I don't think anyone is doubting this bit. It's the overwhelming current trend of forcing everything onto the blockchain, whether it improves the application or not. The interesting questions (frequently not answered) revolve around why one should do this.
There are good arguments for centralization, too. Not saying the blockchain argument win out. Just saying, the benefits are the same as for every other blockchain application, and well understood.
Surely a blockchain would only change that to anybody being able to track your travels, though?
1) open source hailing app or dispatcher 2) rating system with identities
For the hailing or dispatcher, we don't need bitcoin but it may be useful if there is abuse to impose some cost on hails (e.g., uber employees generating fake hails, haha).
For the rating and identity, we can use a GPG web of trust.
Admittedly, the article is light on technical details, but if you understand how blockchain technology works, you could easily deduce that a likely use case for this would be a service for connecting drivers to passengers without a middleman and without the regulatory burden that often prevents these services from proliferating in certain markets. A more fleshed out general case of a decentralized marketplace is OpenBazaar, which creates a marketplace with no middleman that can't be taken down.
There is no reason a blockchain based app couldn't be implemented. Whether this guy is actually on the right path or not is hard to say based on the article.