I suggest you take a look at the perma.cc initative. There's no reason to reinvent the robust shortlink, and you won't be obligated to run this system forever.
"Once a link is vested, it will never be deleted unless we receive a legitimate legal order to take it down."
That seems odd to me. Could they simply publish their mappings periodically to avoid being able to fully comply with takedowns?
I mean, the article using the link is already published, what legitimate reason could there ever be to destroy a link in it? What would these people issuing the hypothetical takedown do if the author of the paper had just included the link itself?
Perma.cc appears to be targeted at links in published journals:
After the paper has been submitted to a journal, the journal staff checks that the provided Perma.cc link actually represents the cited material. If it does, the staff “vests” the link and it is forever preserved. Links that are not “vested” will be preserved for two years, at which point the author will have the option to renew the link for another two years.
I'm sure they'll be willing to work with other publishers as well. From talking to the project manager, they don't seem to have any goal of restricting this to academic publications.
From what I understand, all the member libraries are storing some part of the cache. They might not all have a complete copy, but the details aren't clear yet.
