http://archiveteam.org/index.php?title=INTERNETARCHIVE.BAK
The Archive Team is currently discussing ways to back up the Internet Archive. Fascinating discussion over here as well: Talk:INTERNETARCHIVE.BAK - Archiveteam
Currently they are looking at Storj, Git-annex, and Torrents to back up large files.
This is very interesting; should tell them about MaidSafe’s SAFE Network!, Why not? 
Excited to be able to say “Yeah, just throw it on Safe – it can handle it!”
This is from the Internet Archive, When I first saw this thread, I thought the title said Internet Archive;
I was thinking about Wayback Machine the other day and how safe net could help with this.
As files are made up of immutable data chunks, presumably they could live on long after files associated with them have been deleted. For public files, this made me think - can we store the full history of a file, date stamped and automatically archived?
This would mean that we could have version control built automatically into every file. If you set a date filter, it could mean that the entire network could be referred to in the state maintained at the time - that would be incredibly powerful for historians, never mind being convenient for end users to retrieve old file versions. Obviously incremental file backups would be an easy win too.
Ofc, privacy and security would need to be considered, but public data could automatically behave in this way. Private data could too, if desired.
It is my understanding that this is exactly how SAFE Network works, so implementing the features you describe is pretty much just about writing filters and UI.