Opinion for the Court filed by Senior Circuit Judge WILLIAMS.
STEPHEN F. WILLIAMS, Senior Circuit Judge:The Telecommunications Act of 1996, Pub. L. No. 104-104, 110 Stat. 56 (codified at 47 U.S.C. § 151 (2000)), requires “incumbent” Local Exchange Carriers (“ILECs”) — the Bell Operating Companies and their successors, inheritors of AT&T’s local exchange facilities and services — to lease unbundled network elements (“UNEs”) to their competitors, the competitive Local Exchange Carriers (“CLECs”). See § 251(c)(3) of the Act, 47 U.S.C. § 251(c)(3). The object is to enable CLECs to provide telecommunications services in competition with the ILECs. Petitioner Competitive Telecommunications Association (“CompTel”) is composed of CLECs, many of whom — perhaps all — are also interexchange carriers (“IXCs”). CompTel seeks review of two interim Federal Communications Commission orders, In re Implementation of the Local Competition Provisions of the Telecommunications Act of 1996, Supplemental Order, 15 FCC Red 1760, 1999 WL 1065185 (1999) (“Supplemental Order”), and In re Implementation of the Local Competition Provisions of the Telecommunications Act of 1996, Supplemental Order Clarification, 15 FCC Red 9587, 2000 WL 713746 (2000) (“Clarification”), which impose some limits on CompTel’s members’ access to certain UNEs.
Specifically, the orders address CLECs’ access to a combination of UNEs known as the enhanced extended link (“EEL”). EELs consist of unbundled loops and transport network elements. Clarification, 15 FCC Rcd 9587 at ¶ 2. A loop is a telephone line that runs from the customer’s premises to the ILEC “end office,” which houses switches used to route calls to their destination. A transport then takes the traffic to the IXC or CLEC office, which will route the call to its final *11destination. Leasing this combination of facilities enables new entrants to compete without building their own local loops and transport facilities. And it is especially desirable for them to acquire EELs as UNEs because as such they are priced under a formula of the Commission’s known as “total-element long run incremental cost,” or “TELRIC.” By contrast, the same functions are more costly if they are purchased as a part of the ILEC’s tariffed services (evidently under mandates imposed by the Commission pursuant to § 201), and known in this guise as Special Access services.
An EEL is useful both for the provision of long distance and local service, and the Commission here sought to channel CLECs’ use of EELs toward local service. In the Supplemental Order it limited access to firms who would use EELs to provide “a significant amount of local exchange service.” Supplemental Order, 15 FCC Red 1760 at ¶ 2. In the Clarification it refined this concept and embodied it in numerically defined safe harbors. Clarification, 15 FCC Red 9587 at ¶ 22.
CompTel contests the restriction favoring provision of local service, stating that the 1996 Act does not allow the FCC to make that sort of distinction (referred to as a use or a service-by-service restriction). It further argues that none of the FCC’s justifications for the interim rules makes it acceptable. Finally, it argues that the safe harbor provisions of the order are arbitrary and capricious, mainly asserting that they impose tracking burdens that are difficult or impossible for the CLECs to fulfill and that they impose needless restrictions against commingling of local and long distance traffic. We first address the timeliness of this appeal. Once having found jurisdiction, however, we are unpersuaded by CompTel’s merits claims.
* * *
A petition for judicial review of a final order of the FCC must be filed “within 60 days after its entry.” 28 U.S.C. § 2344 (2000); see also 47 U.S.C. § 402(a). In this case, the petition for review was filed within 60 days of the Clarification but not within 60 days of the Supplemental Order. Respondent argues that it is timely only as to claims that arose from the Clarification, not as to ones essentially aimed at the Supplemental Order.
But the Clarification radically changed the Supplemental Order in a way we have not yet mentioned. In the Supplemental Order the Commission said that it would issue a final decision on the EELs restriction in the Fourth Further Notice of Proposed Rulemaking (“FNPRM”), which notice “will occur on or before June 30, 2000.” Supplemental Order, 15 FCC Red 1760 at ¶ 2 (emphasis added). In its Clarification, the FCC freed itself of this deadline, continuing to state that the order would last until the Fourth FNPRM but giving no time period in which that would occur. Clarification, 15 FCC Rcd 9587, at ¶ ¶ 1, 35.
We hold that this extension newly aggrieved CompTel and thus made its petition timely. See Sam Rayburn Dam Elec. Coop. v. Fed. Power Comm’n, 515 F.2d 998, 1007 (D.C.Cir.1975) (stating that where a party was “aggrieved” by a later interpretation of a rule that it could not have reasonably anticipated, the time limit starts to run at the later event). Cases have held that extension of a temporary order may entitle the parties to seek judicial review of the order. See Public Citizen v. Nuclear Regulatory Comm’n, 901 F.2d 147, 151 (D.C.Cir.1990) (holding that where a temporary order is later made permanent, the permanent order may be challenged); Illinois Cent. Gulf R.R. v. *12Interstate Commerce Comm’n, 720 F.2d 958, 961 (7th Cir.1983) (holding that a temporary order with no fixed time period had lasted so long as to make judicial review timely).
Here the initial order appeared to present CompTel with the prospect of enduring an interim order for a period of no more than six months. After six months, another order was issued extending the time period and setting no ultimate date of resolution. The rules have now been in place for over two and a half years. If we were to hold that such a petition is not timely, a party that found an order not worth litigating because of its apparently short term would have to sue for fear that it might drag on indefinitely. And we might also create a temptation for the FCC to let its deadlines slip. Thus we find the petition for review timely.
CompTel finds in the language of the 1996 Act a bar on the Commission’s making a service-by-service distinction in deciding under what circumstances an ILEC is required to lease UNEs. We review, of course, under the usual standard of Chevron U.S.A. Inc. v. Natural Resources Defense Council, 467 U.S. 837, 842-43, 104 S.Ct. 2778, 2781-82, 81 L.Ed.2d 694 (1984), under which the court must give effect to the clearly expressed intent of Congress, and must if the statute is ambiguous defer to any reasonable construction by the agency. The parties appear to assume that neither the Act as a whole nor § 251(d)(2)(B) in particular requires that impairment findings be service-by-service or that the UNE mandates be confined to services as to which such a finding has been made; accordingly we do not address those issues, but rather try only to answer the question of whether the Act bars such service-by-service distinctions.
The first source CompTel points to as a basis for barring a service-by-service distinction- — in other words, requiring that any UNE be mandatory for all telecommunications services, once it is found appropriate for one telecommunications service — is § 251(c)(3). It says that incumbent local exchange carriers have
[t]he duty to provide, to any requesting telecommunications carrier for the provision of a telecommunications service, nondiscriminatory access to network elements on an unbundled basis at any technically feasible point....
47 U.S.C. § 251(c)(3).
We do not see that the phrase “for the provision of a telecommunications service” helps CompTel. If Congress had spoken of “the provision of any telecommunications service,” the language might conceivably be taken to suggest that once an element was ordered to be made available for one telecommunications service, it must be made available for all. But the vaguer phrasing chosen by Congress does not lend itself even to that suggestion.
CompTel also sees hope in § 251(d)(2)(B), which states that
[i]n determining what network elements should be made available for purposes of subsection (c)(3) of this section, the Commission shall consider, at a minimum, whether— ...
(B) the failure to provide access to such network elements would impair the ability of the telecommunications carrier seeking access to provide the services that it seeks to offer.
47 U.S.C. § 251(d)(2) (emphasis added). Far from helping CompTel’s cause, this passage seems to cut strongly against it. By referring to the “services that [the requesting carrier] seeks to offer,” it *13seems to invite an inquiry that is specific to particular carriers and services. That the Commission has never gone down the path of carrier-by-carrier UNE analysis provides no basis that we can see for making it forego service-by-service limitations.
In United States Telecom Association v. Federal Communications Commission, 290 F.3d 415 (D.C.Cir.2002) (“USTA”), we clearly found in the FCC an authority to make distinctions that were based on regional differences or on customer markets. See id. at 422-23. If these are permissible, it is hard to understand why the Act would not allow restrictions keyed to a specific “service” of the requesting carriers. The Clarification offers the essential and compelling reasoning:
Although ambiguous, that language [§ 251(d)(2)] is reasonably construed to mean that we may consider the markets in which a competitor “seeks to offer” services and, at an appropriate level of generality, ground the unbundling obligation on the competitor’s entry into those markets in which denial of the requested elements would in fact impair the competitor’s ability to offer services.
Clarification, 15 FCC Red 9587 at ¶ 15. CompTel’s theory, by contrast, would mean that once the Commission found a single purpose as to which an “element” met the impairment standard, no matter how limited, it would be forced to mandate provision of the element for all, no matter how little potential impairment was involved in the remainder of the telecommunications field. CompTel never explains what logic could have persuaded Congress to lock the Commission into such a scheme.
CompTel also claims support in the Act’s definition of a “network element” as “a facility or equipment used in the provision of a telecommunications service.” See 47 U.S.C. § 153(29). We cannot discern a bar on service-by-serviee mandates in this bland phrasing.
Finally, CompTel argues that past orders of the FCC make it clear that the FCC understood the statute not to allow it to make a service-by-service distinction. CompTel points to orders issued by the FCC both before and after the court’s decision in AT&T Corp. v. Iowa Utilities Board, 525 U.S. 366, 119 S.Ct. 721, 142 L.Ed.2d 835 (1999). See In re Implementation of the Local Competition Provisions in the Telecommunications Act of 1996, First Report and Order, 11 FCC Rcd 15499 at ¶ ¶ 264, 356, 358, 1996 WL 452885 (1996) (“Local Competition Order”); In re Implementation of the Local Competition Provisions of the Telecommunications Act of 1996, Third Report and Order and Fourth Further Notice of Proposed Rule-making, 15 FCC Rcd 3696 at ¶ 484, 1999 WL 1008985 (1999) (“UNE Remand Order”). In the Clarification the FCC indeed acknowledged that
[b]efore the Supreme Court issued its decision in loica Utilities Board, we sometimes approached an incumbent’s obligation to unbundle network elements as though it were an all-or-nothing proposition, suggesting that, if a competitor were entitled to obtain access to an element for one purpose, it was generally also entitled to obtain access to that element for wholly different purposes as well.
See Clarification, 15 FCC Rcd 9587 at ¶ 12. The FCC further acknowledged that it had not properly focused on the relationship between that issue and the impair standard, and that the court’s opinion in Iowa Utilities Board made it appropriate to “revisit the issue.” Id. The Commission is clearly correct that Iowa Utilities Board required it to limit its former all-encompassing interpretation of the necessary and impair language of 47 U.S.C. *14§ 251(d)(2). Iowa Utils. Bd., 525 U.S. at 388-90, 119 S.Ct. at 734-35; see also USTA, 290 F.3d at 418-19.
CompTel points out that the FCC’s statement in the UNE Remand Order followed Iowa Utilities Board; CompTel reads the statement as a reaffirmation of a prohibition on use restrictions. But the relevant section plainly does no more than accurately recount past Commission conclusions, repeatedly saying, “the Commission found ...,” and closing with the observation that its rule against usage restrictions “was not challenged in court by any party.” UNE Remand Order, 15 FCC Red 3696 at ¶ 484. Other provisions of the order made it clear that the FCC did not consider the issue at all settled. See, e.g., id. at ¶ ¶ 489, 493-96. The FCC is obviously entitled- — indeed required — to reconsider orders that rest on faulty readings of a statute. That is all it has done here.
CompTel next argues that even if the Commission is authorized to place service-by-service restrictions on the use of UNEs, it acted arbitrarily and capriciously because its justifications for doing so here were unreasonable. As we said above, all hands assume that the Commission can mandate provision of a UNE for all telecommunications services even though its finding of § 251(d)(2)(B) impairment is based on only one service. But it is far from obvious to us that the FCC has the power, without an impairment finding as to nonlocal services, to require that ILECs provide EELs for such services on an unbundled basis. Here there is no finding by the Commission that lack of access to EELs would “impair” CLECs’ ability to provide long distance or exchange access service. Rather the FCC observed in the Clarification that it had the power to inquire into whether there was any such impairment, Clarification, 15 FCC Red 9587 at ¶ 15, but it then explicitly declined to reach that question without a chance to “gather evidence,” id. at ¶ ¶ 16-17. Comp-Tel makes no argument that the Commission should have (or even could have!) made any finding without further evidence. As both parties assume, however, that the Commission can require a network element to be unbundled for all services when it only finds it to be impaired as to one service, we accept the assumption arguen-do and address the question of whether it exercised reasonable judgment in withholding such a requirement here.
The FCC gives in essence two justifications for its interim restrictions on the unbundling of EELs. First, it argues that the rule is necessary to avoid disruption of its reform of access charge policies and of the implicit subsidies for universal service that remain embedded in access charges. It also reasons that its restrictions are needed to promote faeilities-based competition.
Avoidance of market disruption pending broader reforms is, of course, a standard and accepted justification for a temporary rule. See, e.g., MCI Telecommunications Corp. v. Fed. Communications Comm’n, 750 F.2d 135, 141 (D.C.Cir.1984); see also ACS of Anchorage, Inc. v. Fed. Communications Comm’n, 290 F.3d 403, 410 (D.C.Cir.2002). Here the broader reform involves an attempt to make access charges more truly cost based and to correct the divergences from cost that have been mandated in the interest of implicitly subsidizing some users at the expense of others. See In re Access Charge Reform, First Report and Order, 12 FCC Rcd 15982 at ¶ ¶ 42, 44, 1997 WL 268841 (1997) (“Access Charge Reform Order”). Both of the orders that the FCC cites as part of its “access charge reform policies” are steps *15on the way to eliminating implicit subsidies. Id.; In re Access Charge Reform, Sixth Report and Order in CC Docket Nos. 96-262 and 94-1, Report and Order in CC Docket No. 99-249, Eleventh Report and Order in CC Docket No. 96-45, 15 FCC Red 12962 at ¶¶ 24-25, 2000 WL 1760912 (2000) (“CALLS Order”).
The link between access charge reform and unrestricted access to EELs is that without the restrictions competitive carriers would use EELs as an alternative to switched access services and thus avoid paying the implicit subsidy, undermining the subsidy scheme. Supplemental Order, 15 FCC Red 1760 at ¶ 7.
CompTel does not appear to deny either that the Commission can justify a policy by reference to the purposes of avoiding disruption pending a broader reform, or that the EEL restriction would disrupt implicit subsidies contained in access charges. Rather, it asserts that the access reform is actually complete rather than pending, so that the Commission’s justification is pure pretext.
The first variation of this claim is an argument that 47 U.S.C. § 254(a) requires elimination of implicit subsidies within 15 months of adoption of the Act, i.e., by May 1997; thus, apparently, any residual implicit subsidies are illegitimate and not a permissible basis for the Commission’s orders. This is a misreading of § 254(a). The section requires that the Commission start a proceeding to assess recommendations of a Federal-State Joint Board on the subject and complete the proceeding within-15 months of the date of enactment, setting a timetable for the reform. 47 U.S.C. § 254(a). But there is no time limit on realization of the reform. And the petitioner offers no evidence to show why the proposed timetable given by the CALLS Order is unreasonable.
Second, CompTel argues that the CALLS Order, issued two days prior to the Supplemental Order, immediately eliminated implicit subsidies for universal service, replacing them with an explicit mechanism of support. Since the FCC has totally removed implicit subsidies, Comp-Tel claims there are none left to protect. In support CompTel cites ¶ 202 and other sections of the CALLS Order, arguing that the FCC estimated the cost of universal support to be $650 million per year and immediately created an explicit support mechanism of exactly that amount. See CALLS Order, 15 FCC Red 12962 at ¶ ¶ 30, 32, 186, 196, 201, 202, 226. But a close look at the CALLS Order makes clear that implicit subsidies are phased out over time.
The CALLS Order identifies a number of elements in access charges that have been sources of revenue for implicit subsidies. See id. at ¶ 23. It names (1) the multi-line business Presubscribed Interexchange Carrier Charge (“PICC”), which is a flat per-hne charge imposed by a ILEC on an end user’s IXC, see id. at ¶ ¶ 2, 19; (2) Carrier Common Line (“CCL”) rates (per-minute charges assessed on the end user’s IXC whenever the end user placed an interstate long-distance call), see id. at ¶ 18, which have per-minute charges for a service that has a fixed cost; and (3) geographic averaging, which suppresses geographically based cost differentials and subsidizes users in the more costly low-density areas. Id. at ¶ 23.
The CALLS Order’s treatment of these charges makes us confident that it did not stop the subsidies on a dime. While multi-line business PICCs are higher than would be appropriate (vis-a-vis cost) given the caps on subscriber line charges for residential and single-line businesses, id. at ¶ 23, the CALLS Order phases them out over a five-year period. See id. at ¶ 110 *16(“For example, we estimate that by July 2004, the multi-line business PICC will be eliminated under ... the proposal for all BOCs except for BellSouth, which would have a multi-line business PICC at that time of approximately $0.20 per line.”) If this were not strong enough evidence, ¶ 111 specifically recognizes that the existence of this rate in some areas “may constitute an implicit non-portable subsidy.” Id. at ¶ 111. It then goes on to justify this subsidy as an interim mechanism that provides a reasonable solution to the tension between protecting universal service and providing cost based rates. Id.
Similarly the Commission provides for gradual phase-out of the CCL charges, see id. at ¶ 144 (“Upon the earlier of the elimination of the CCL charges or June 30, 2004”), speaks of “reducing” averaging, id. at ¶ 26, and indicates that deaveraging will proceed under “limited conditions and safeguards,” id. at ¶ 115.
Thus, in three major areas, subsidies are gradually pared back over time. As the implicit subsidies for universal service have not been eliminated, the FCC has an interest in preserving the process of their gradual reduction and extinction.
Finally, CompTel attacks the Commission’s explanation of the orders as an effort to protect facilities-based access providers. Invoking this interest, the Commission said that immediate provision of special access as a UNE “could undercut the market position of many facilities-based competitive access providers,” a form of access that originated in the mid-1980s and “is a mature source of competition in telecommunications markets.” Clarification, 15 FCC Red 9587 at ¶ 18. CompTel argues that favoring facilities-based competition is not a valid policy goal.
But the Supreme Court’s discussion of the incentive effects of TELRIC in Verizon Communications, Inc. v. Federal Communications Commission, — U.S. -, 122 S.Ct. 1646, 152 L.Ed.2d 701 (2002), would be meaningless if the Court had not understood the Act to manifest a preference for facilities-based competition. Indeed, the Court puts it expressly, characterizing the ILECs’ critique of TELRIC as a claim that its “result will be, not competition, but a sort of parasitic free-riding, leaving TELRIC incapable of stimulating the facilities-based competition intended by Congress.” Id. at 1669. Obviously accepting the ILECs’ view that Congress preferred “facilities-based competition” over “parasitic free-riding,” the Court went on to decide that the FCC could reasonably conclude that TELRIC would not create the perverse incentives the ILECs claimed. Id. at 1669-78. See also our opinion in USTA, 290 F.3d at 424-25.
In this case, the FCC has only issued an interim rule while it further studies the issues to determine what rule will best promote facilities-based competition. Certainly when combined with the other rationales in this case this constitutes adequate justification for the rule.
Finally, CompTel makes two arguments that the safe harbor provisions are arbitrary and capricious. First, it says that CLECs don’t have the information needed to comply with them, and that the difficulties of obtaining such information will in most cases outweigh the benefits of the EELs. Second, it argues that the commingling restrictions will deny its members a substantial market benefit of the EELs and, like the information requirements, are unnecessary.
*17Expressed in its most plausible form, CompTel’s argument about the safe harbors’ information demands — going simply to the usage patterns for the facilities — is that the data are not information that the companies normally obtain, that obtaining the information and insuring its accuracy would be difficult and costly, and that in most instances the costs of complying outweigh the benefits of the EELs.
The difficulty with this argument is that it suggests no alternative way for ensuring that EELs are only provided where they are used for “a significant amount of local exchange service.” Thus, if we were to find that the safe harbor provisions were too demanding on carriers, we would essentially be deciding that the FCC simply could not implement its use restriction because it is not administratively feasible. But it is plain that supplying the information is feasible, as the FCC has produced evidence that some carriers are taking advantage of the safe harbors.
CompTel’s second target, the Commission’s anti-commingling rule, essentially “does not allow loop-transport combinations [taken as UNEs] to be connected to the incumbent LEC’s tariffed services.” Clarification, 15 FCC Rcd 9587 at ¶22. CompTel and intervenors argue that it is much more cost effective to use the same transport facility for both local and long distance than to have two separate facilities, each carrying a smaller amount of traffic. They further argue that there is no reason not to allow commingling, as it is impossible for parties to use commingling to eviscerate the local use requirement. The FCC, on the contrary, argues that a restriction against commingling is the only way to prevent carriers from using these units “solely or primarily to bypass special access services.” Id. at ¶ 28. We find that CompTel has not met its burden of showing that the restriction is arbitrary and capricious.
One of the difficulties of this question is the way in which it is raised and addressed in the briefs. In their opening briefs both CompTel and intervenors argued that the restriction on commingling was oppressive, and they stated in a cursory manner (CompTel devoting all of one page to the issue) that it was not needed to carry out the FCC’s goal of preventing special access bypass. The FCC responded in an equally cursory way, saying that allowing commingling would allow carriers to avoid the requirement that each customer be provided a significant amount of local exchange service. In their reply briefs, both petitioner and intervenors explained in more detail exactly why allowing commingling would not result in “gaming” of the system. As we explain below, we find in the record indications of why the Commission might have fears about commingling' — even under the conditions suggested by the petitioner and intervenors. The FCC did not raise these reasons in its brief, but we cannot fault it for this omission given the sketchy character of Comp-Tel’s and the intervenors’ initial arguments. “[T]he Commission cannot be asked to make silk purse responses to sow’s ear arguments.” City of Vernon v. FERC, 845 F.2d 1042, 1047 (D.C.Cir.1988); see also McBride v. Merrell Dow and Pharms., Inc., 800 F.2d 1208, 1210 (D.C.Cir.1986) (“We generally will not entertain arguments omitted from an appellant’s opening brief and raised initially in his reply brief.”)
There are, as we said above, two more complex reasons why gaming might occur. The first appears to be that commingling will allow the entire base of the loop or “channel termination” portion of special access circuits to be converted into unbundled loops. The reason is that there are *18no use restrictions on unbundled loops, and therefore allowing loops to be freely-connected to special access services would allow loops that provide no local services to be unbundled and then merely attached to special access transports. See May 11, 2000 letter from MCIWorldCom to FCC at pp. 2-3 (“WorldCom Letter”), Joint Appendix (“J.A.”) at 418-19. The second concern appears to depend on possible erosion of the line between UNEs and certain non-UNE capabilities that an ILEC, seeking authority to offer long distance service in the region where it provides local services, is required to offer under the so-called “competitive checklist.” See 47 U.S.C. § 271(c)(2)(B); UNE Remand Order, 15 FCC Rcd 3696 at ¶ 468; WorldCom Letter at 3-5, J.A. at 419-21. The idea is evidently that through the interaction between § 251 and § 271, free “mixing and matching” in the present context would lead ultimately to CLECs’ securing UNE rates for these competitive checklist capabilities even when the Commission had found that lack of the capabilities would not meet the “impair” standard for UNEs under § 251(d)(2). WorldCom Letter at 3-5, J.A. 419-21.
As we have said, the FCC did not present this argument or point to a place in the record where it had made it. We know of it from a WorldCom letter to the FCC articulating it and a possible solution that WorldCom said would still allow commingling in the case at hand. See WorldCom Letter at 2-5, J.A. 418-21. Because neither CompTel nor intervenors dealt with these more difficult questions in their briefs, we are in no position to assess the feasibility of the solution, or to decide whether the FCC might have solved the problem in a different and less burdensome way. But on the present record we are plainly unable to say that the restriction on commingling is arbitrary and capricious.
The intervenors raise a third issue (collocation) regarding the safe harbor provisions, but since this was not raised by the principal parties, the intervenors are pro-eedurally barred from arguing it. See United States Tel. Ass’n v. Fed. Communications Comm’n, 188 F.3d 521 (D.C.Cir.1999).
For the reasons stated above, the petition for review is
Denied.