This is harder than you might expect. Publishing these files is always risky because sites can serve you fingerprinting data, like some hidden HTML tag containing your IP and other identifiers.
For a historical archive, the issue with this is that it could be difficult to ensure that the data being sent from users' devices wasn't modified in some way, leading to an inaccurate archival copy.
Cross-reference. When a site is archived by one client (who visited it directly), request a couple other clients to archive it (who didn’t visit it directly, instead chosen at random, to ensure the same user isn’t controlling all clients).