neither does arweave lol, they just think it will
what would solve the problem is more work on dynamic, decentralised replication algorithms... which don't exist yet, ipfs is slow and limited, the only successful protocol that kinda almost works is bittorrent which really says a lot, i hope blossom comes to be the solution for this...
the archival problem is just a matter of being able to get paid to keep the giant stash
exactly, it's about incentives at the edges
Archives don't have to be decentralized.
It would be enough to have some centralized archives that fetch data from other relays and expressly aim at being archives that preserve posts for as long as possible (unlike most relays, which have no express aim or policy).
The amount of data would, likely, not even be that large (it's just text), especially if you filter out spam (which you may or may not want to do). The price of storage wouldn't be a significant issue.
nostr:npub1lunaq893u4hmtpvqxpk8hfmtkqmm7ggutdtnc4hyuux2skr4ttcqr827lj has ideas about this too
