[mikeash]

The Shuttle's cheapness was predicated on a high launch rate. When the high launch rate wasn't achievable, it became pretty expensive. That's only to be expected, of course. A 747 is an economical way to fling a bunch of people and cargo across the Pacific if you do it once a day, but it's terribly expensive if you only fly it a couple of times a year. The Shuttle was decent enough conceptually, but it pushed the envelope in too many places and that hurt reusability a lot.

I agree that SpaceX's setup is more promising. The fact that they're slightly modifying an existing expendable rocket that's already cost-competitive means that any reuse is just a bonus. They can throw away a bunch of boosters as they work on the problem, which wasn't an option with the shuttle at $2 billion apiece just to construct them.