I'm not questioning that there are great engineers working on crypto, there are amazing engineers working on ad tech and high frequency trading too. Just like those examples I'd assume it's mostly because it's an interesting engineering problem and there's a ton of money to be made.
The only exciting potential use case of crypto that I've seen so far are DAOs, but I'm not convinced that there's any reason why they need to be on a blockchain instead of postgres or sqlite.
Many of us have a practical attitude which is effectively that the technology has massive potential, and at the same time, that it's probably overvalued for the current state of technology.
Of course, overvalued is subjective. It could all still amount to nothing, but it could also replace tons of legacy/traditional infrastructure and systems, and/or create a new global financial layer based on currencies whose supplies can't be manipulated by governments. The valuation is of course massively coupled to the perceived potential and the varying levels of conviction speculators have in it, which incentivizes a lot of disingenuous behaviours.