Perma! Making (some of) the Internet forever

I heard a piece this morning on Marketplace Tech Report about Perma, which is a program started by a group of libraries in order to make online materials cited in scholarly journals permanent. The members of Perma are trying to solve a real problem: scholarly articles regularly cite links, and there's no guarantee that those links won't become obsolete. This problem, called link rot, is pervasive; according to Perma, 70% of the links in articles published in legal journals between 1999 and 2011 direct users to irrelevant content. Perma creates a permalink to an archived version of the original page.

Perma has some gatekeeping built in: once you register a link that you want to preserve, the scholarly journal that you are citing it in has to verify (or vest) the link to say that, yes, this is being cited in a scholarly source. The journals themselves have to be verified (or sponsored) by an institution. So, you can't just register every website that you think it would be great to have a copy of (NOT that I was tempted to do that. Certainly not).

Perma sounds like a great idea, but I have some questions...

1) If I'm understanding Perma's mission, a lot of the links that get perma'd are going to be to scholarly journal articles. If that's the case, how do we handle paywalls? I access journal articles through my library's subscription, and the link is generated through our proxy server. How do users from different institutions (or those who aren't affiliated with institutions) access subscription content archived through Perma?

2) Likewise, how does Perma overlap with other archiving programs like LOCKSS (Lots of Copies Keeps Stuff Safe)?

3) How will this impact open access journals that may not have institutional backing? Can they also be sponsoring journals so that the links in their articles are archived?

4) This only solves part of the problem for researchers using online sources. Wouldn't it be great if we could also solve the problem of archiving content that isn't necessarily cited in scholarly journal articles (or that future scholars might want to cite, but wouldn't have access to because it will have disappeared)? Too bad the NSA's new data farm in Utah can't be put to this use...