> I personally haven't seen any problem solved by the blockchain thus far.
I don't know that it counts since it's not a "blockchain" in the sense most people describe it, but Git is sort of "blockchain adjacent." It uses a Merkle tree to maintain a crytographically signed quasi-immutable ledger of transactions.
While it's probably not a "blockchain", I don't think that anyone disputes that Git (and its blockchain-like feature) is useful.
I suspect that this is going to be the "good" that comes from blockchain [1] and its ilk; a lot of interesting uses are going to pop up using blockchain concepts, and it will be cannibalized into something that is indisputably useful.
I agree that there are a lot of crypto idiots who act like all problems will be solved by Bitcoin or Ethereum, and I do think they give the legitimate research being done on blockchain (e.g. the Ouroboros protocol) a bad name.
[1] I know Git predates blockchain.