>very large contracts which result in absolutely nothing
That's a moderately ironic statement, given the fact that SpaceX wouldn't be doing anything remotely close to what they're doing today without the very large contract they got from NASA.
Actually, no irony at all here. The large contract that SpaceX (and Orbital Sciences) received were for shipping cargo to the space station -- cash-on-delivery. No delivery, no cash. This is profoundly different from the business-as-usual NASA contract which is: give a contractor a billion dollars. Contractor produces a bunch of powerpoint slides. Give the contractor another billion dollars. Contractor goes a further billion dollars over budget, and produces a bunch of powerpoint slides. Rinse & Repeat.
This is how NASA was able to spend over 30 years and $30B trying and failing to develop a new orbital vehicle, where SpaceX was able to do it for under $400M.
Without NASA contracts (to be noted: fixed price contracts predicated on delivery of goods, for the most part) SpaceX would have much less cash on hand and their pace of R&D would be much slowed. But they would still exist and still be pushing the state of the art, just at a slightly slower pace. They have one of the most competitive orbital launchers on the market, they have a ton of commercial business already on the docket, and the next 3 SpaceX launches are, in fact, non-NASA commercial flights (a Canadian weather satellite, a commsat for servicing East Asia and Oceania, and several Orbcomm commsats).
That's a moderately ironic statement, given the fact that SpaceX wouldn't be doing anything remotely close to what they're doing today without the very large contract they got from NASA.