I concur--it's a flawed analogy on a couple levels (but still a good conversation stimulator) but I believe one of the hardest things to manage in a start-up is that your success isn't binary.

Yes, there are a great many start-ups that fail to get out of the gate and a small, small number that take off so quickly that success is almost a guarantee. But if you're in the muddy middle the decision to grind it out or try a new approach has to be a difficult one, and is usually a low-data decision. And platitudes are little help here with "it takes three years to create an overnight success" on one side and "fail fast" on the other.

There's also some danger of being stuck in a state of mediocrity.