> This is the problem, and as you implicate,
  > the exact sort of thing we need to stop doing.

You do see the absurdity here, right? Banking on substantially above-average developers doesn't work at scale, by definition.

Sure. So long as you have a full proof way of identifying the best developers.

Hint: the best developers aren't the ones who can recite the API off hand.

You have to be willing to fire them if it turns out they're they suck.

Average != suck.

Anyone who thinks they are going to do nothing but hire above-average programmers at any scale is downright naive. Design your system around an "average" employee, and you'll be far happier.

You can do a hell of a lot with a 10-15 person team. Unless you're consulting your company can grow much faster than your team.

Fool. Fool proof. Not full.

Thanks, you really honed in on that mistake. <runs away>

> Sure. So long as you have a full proof way of identifying the best developers.

It's easier to grow them.