This isnt my view. I dont think any crypto project delivers -- and this is an extremely strong claim requiring only a single counter-example. I'd love to hear the counter-example.

I dont think a distributed immutable ledger whose state changes one when everyone recomputes it is a good technology for almost anything. Think of this from a devops/data-eng perspective... what, in your infrastructure, do you want to work that way?

In practice, we are always able to assume a non-trivial amount of trust which would eliminate almost all of that excessive computation. OR, if not, coordinate in such a way that we can acquire non-trivial amounts (eg., first exchanging keys).

This technology is an engineering revolution in the minute "data-engineering cyptographic problem space". That might be a business somewhere. I dont think it has almost anything to do with any marketable proposition.

I cannot think of any single business case thus far offered where removing the blockchain doesnt produce a better product.