There's so much opportunity for Orwellian abuse in this new archival model.
Sure there's archive.org - but in the old model, you'd say something in print, copy it a million or so times, then distribute it on paper around the world.
If an evildoer wanted to change a claim from say, 20 years ago, they'd have to create a forgery that looks like it's 20 years old, then secretly swap out all the archival copies in libraries and get large-scale mass obedience from librarians or manually make the swap at each library individually.
These days, you need to add an exclusion to the archive.org bot in robots.txt and then flip a few bits. It's a different ballgame.
In the traditional model, someone could find say, an original copy of the 20 year old newspaper and present that as evidence. Forging that document would be sufficiently hard for a single individual. But in the new model, claiming something is an "original copy" of a web page can be easily disputed because such a forgery can be trivially crafted.
Think about the implications in places with large civil unrest, very little free press, and mostly state-owned media. What's the solution for those societies?
It's a problem.