* Decentralized digital Asset ownership verification. Again, imagine facebook without a facebook. Or imagine purchase a music or software token that you could use on any provider, and never expired. Sign up for a new streaming service and take all your songs/videos/games with you from the old one.
* All the legal structures of a company, purchased off the shelf, handling all the vagaries of human resources and accounting, but without managers or payroll accountants. Just remote teammates, who may never have worked before, but know they will get paid a salary, receive evaluations and promotions from their peers.
These are just toy examples, I'm sure if I took more than a few seconds, I could give you much better ones. The problem with blockchain isn't finding applications - it's finding monetizable ones.
All the examples above would make amazing open source projects, but IMHO would be very difficult to generate revenue without compromising their decentralized nature, by inserting oneself into the transactions to take a cut. Centralized control of an application is almost a requirement for monetization, and, blockchains, are, by definition, intended to circumvent centralization.
Moreover, trust and verification aren't new problems, and we already have plenty of centralized solutions for these problems. So, blockchain applications both have to be better than what was there before, and work around the decentralized nature of the system.
Consequently, people trying to make a living on the blockchain mostly resort to monetizing the networks themselves - in other words, developing cryptocurrency-like products of some form. Is this decentralized? Not entirely, but it's less centralized than what came before. If people disagree with the behaviour oof the ethereum foundation, they can fork the currency. However, this still a long cry from actual decentralization. It is certainly an open question what value 'kindof-sortof' decentralized networks provide.