Wikipedia once had a way to download a big tar(?) archive of their articles, IIRC in wiki syntax. I don't know if this function is still active.

Having a cron job export the source of all nodes readable to Anonymous Monk every day, week or month to an archive might be an idea. That joub could perhaps run on a dedicated server, perhaps on a snapshot of the live database, so it would not put extra load on the normal servers.

Alexander