A blockchain is at the end of the day nothing else than a distributed database where everyone can run nodes. This is a hell of an effort and everyone who is using this 'feature' needs a very good reason and use case which justifies the costs. Every new ICO should answer the question if a distributed BFT database is required. If yes is the benefit big enough? If not then just forget it. Or invest and sell the first mi…
* if you have a technical problem that requires decentralised, cryptographically verified tokens (if it doesn’t need tokens, they shouldn’t be bolted on);
* if the tokens are directly usable on the platform itself;
* if at least a proof-of-concept of the technology verifiably exists;
* it also helps if the idea is even plausible as a business.
Unsurprisingly, most ICOs don’t meet these criteria.
The closest I've seen is Civic, which tried quite hard to meet the first three. However, for the fourth, it's still using a blockchain to solve a problem that only exists if you use a blockchain.