Part of the problem is that archive.org is effectively a single target. The larger they get, the more content they house, the more devastating it would be to have them removed and the more logical this ultra-gun-shy approach to legality is.

Moreover, states and corporations are naturally going from incidentally targeting archived content to systematically targeting it - so that through a flood of incidental and purposeful copyright attacks, archived content may wind-up a very well "pruned" set indeed.

It seems like having a less "monolithic" approach to archiving may be necessary - I'm not sure what that would look like, networks of smaller archivers who try to cover everything? Tor? Freenet?