I don't think the approach taken by schema.org is the best one, far from it. But the Semantic Web has got quagmired so many times, while the rest of the world gets on and does cool stuff. I recently got irritated enough by the goings-on in the HTML5 WG to join [c.f. extensibility] but then realised I couldn't make any difference, certain patterns/arguments are set in stone. Facebook has taken the world by storm with ideas that were in FOAF years ago. Ball dropped.
I used to hold the opinion that the shortest path from A to B was usually the best one. But any path is better than spending more time wading the river.
Also Dan Connolly came out with a shrewd phrase like "so which constraints do we need to relax for this to go viral?"
And Michael's already sorted http://schema.rdfs.org, mappings to RDF schema :)