Having worked at two startups that died either because they could never attract skilled senior engineers or because they lost them, I think the number is quite large.

There're lots of ways for a startup to die and relatively few for it to succeed. The value of having a skilled senior engineer (or "rockstar ninja" as you put it, though that term has connotations that don't really describe the type of early engineers you want) is that they can see many of those failure modes ahead of time and avoid them. And when they do end up in a dead-end, they can quickly backtrack and fix the problem instead of wondering "what do we do now?"

Businesses fail to get traction for product reasons, business fail to grow/thrive because the mis-managed complexity makes changing the system too expensive. Ex: well, i'd love to build you that feature, but then i'd have to change x, y and z and there are no tests for y or z because we thought it would unnecessarily slow us down.. so it will take 3 months instead of the 3 weeks that hotshot rails dev you just hired said it should.

Few companies need the 0.001% that is a "Rock star ninja" such as Larry Wall or Dennis Ritchie.

However, to succeed you do need to be able to hire the 0.5% of developers that is competent and capable of bringing value.