The idea that the blockchain is something that can be used as a database seems to be one of the big Web3 claims. Clearly, it's a thing in which arbitrary data can be stored; but that's not it's primary purpose, and it seems to be an astonishingly-inefficient way to store data.

If someone were to help me understand why it makes sense to use blockchain as a database (or as a webserver), maybe the scales would fall from my eyes, and I'd begin to grok "Web3". But as far as I can see, the only relationship between records in the blockchain is that each record has all earlier records as its parent. You can't establish any other relationship between two records, other than that each one contains a hash of all its parents.

Yes, a ledger has an implied schema. But it's a single table, so it's not much of a big deal, as schemas go. And everyone who wants to use the blockchain as a database is obliged to use the same implied schema. A flat text-file would be more versatile.