There is an interesting pattern I've noticed: there is a whole lot more diversity of "types" in the successful bootstrapped teams I've met (that eschew traditional investment models) than in the VC funded/VC intent teams. Especially if you scan the international scene.

The answers proposed so far are largely consistent, and I think they can be composed into a consensus view. A friend of mine actually has a name for it. He calls it the "Valley Dogma."

I can think of examples of almost every one of the rules being proposed here being broken, and in most cases, the examples that are coming to mind are bootstrapped teams, often not operating in Silicon Valley.

The constraints imposed by the technology set are actually far weaker than those imposed by the success narrative du jour: (YC+idea-->angels+traction-->VCs+growth-->billion-dollar exit).

Which suggests to me that this Valley Dogma consensus answer is not a necessary condition for success in general, but a necessary condition for VC-path/Valley model success. It is a de facto necessary condition for success in general to the extent that angel/VC money is the only money in the game.

So... not a bad idea per se to follow the Valley Dogma answer, but there are many more ways to skin the cat.

I think the paradigm difference can be captured in two alternative axioms:

1. Valley Dogma: Given the success narrative du jour, what does the right team look like?
2. Given a team of talented people who get along, what can they successfully do together?

Even within the Valley Dogma, it is important to keep a couple of formula-break possibilities in mind, that may help you:

1. Given the specific idea, there may be specific pieces of the formula you can break, thereby gaining high leverage
2. You have far more flexibility in the paths you can take to the formula, based on timing and sequencing, jury-rigging stop-gap teams together, cobbling together Frankenstein star talent by mashing up lesser talent and limited time from advisory talent