> True, but it's entirely possible that such a use case will never materialize
Hard disagree. DAOs already exist, for example, and they are widely used even if the phenomenon is unknown to who doesnt have an interest in it or who is not technical enough. They are still way too technical for the average user, and even solely this use case could easily explode the moment everything is simplified and presented to the users with an intuitive and easy to use interface by a new service.
A lot also depends on the infrastructure improving - IPFS is great, for example. But its not as fast as a server, forcing you to store very little data using it. For the moment, any use case that will involve IPFS must fit into those narrow confines. Actually, same goes for most of crypto - Ethereum is a framework that you can literally, actually program on. But you wont be compiling any complex modern app code and serving it from there. One reason why the use cases still revolve around financial things, NFTs and other things that are built on the ledger functionality.
> but it won't be one resembling what anyone thinks
That's entirely possible.
> and won't be sexy.
And that totally depends. Anything that gets mass adoption ends up 'sexy'...