Proof Of Transfer (private)
In my opinion, the largest currently unsolved problem (as of 2023) in decentralized economic filesystems (e.g. crypto file markets) is the strategy-proof retrieval of data.
This is the core principal of usefulness of the Internet, in my opinion - getting data. In fact, that may be the defining functionality - what is the Internet without retrieval? You can't visit a website (retrieve html), consume media/information (download image/audio/video), and most of the rest. (One may still be able to do 1-1 messaging, but it would require an updated trust model).
Thus, I found it surprising that SiaCoin, Storj, and Filecoin all originally opted to waive their protocol-level responsibility of ensuring data retrieval.1 This indicated to me that there was difficulty finding a distributed algorithm with the proper incentives and encryption to create the fundamental transaction guarantee necessary to decentralize the Internet.
I believe I have developed such a distributed algorithm, and I'd like to open-source it someday. But for now I think I'll keep it a little private and see what I can do with it. If you're interested in talking about it, I'm not closed to it - but I'd rather not publish it just yet.
Footnotes
Many hundreds of millions of dollars later, Filecoin did put together a 'retrieval market' algorithm whereby data was transfered bit by bit with corresponding micropayments. This simply pushed the strategy-proofing problem to the last packet, which means that encrypted files are totally irretrievable with a short-term profit-maximizing peer.