I find that silly. I'm not a chef but I can taste food.
Increasingly I have the tendency to believe that all this technical/non-technical is a really unfortunate simplification. I think especially when it comes to a startup tackling really big problems whether you've programmed before might be a rounding error in the equation toward succeeding.
Particularly for sniffing out security problems... you're fighting a many-headed beast, so unless you're quite technical yourself (and know how to exploit the whole OWASP list, for example), I can't imagine how you'd evaluate someone else's statement that "yes, this will be secure"; even if the lead developer can talk in depth about 10 common security holes, what if s/he simply isn't familiar with #11 and #12? Or lacks the creativity to notice how an architecture choice will severely hamper security in the future?
There's non-trivial work involved in evaluating my own work, and every now & again when I step back to view my own approach to a technical problem I change course.
So -- it's certainly possible to evaluate developers without being technical, but you're still forced to trust their diligence and skill quite a bit. "Talking the talk" of doing things right technically isn't very hard (just read a lot of dev blogs... you'll pick it up); actually doing them right with some consistency is a different beast entirely, and not everyone succeeds even with the best of intentions (...but this is much harder to evaluate).
Reminds me of this Joel Spolsky essay: http://www.joelonsoftware.com/articles/fog0000000069.html
The focus being on people does not preclude writing good code.
That's an unfortunate simplification. Without a technical background, you wouldn't know how long it takes to make a basic CRUD app. "3 weeks? Ok I guess. Oh, 3 months actually? Ok I guess." And what about performance, scalability, costs, etc? You'd have no clue.
Also, and perhaps most importantly, when I find a minor bug in my app, I can fix it myself and push an update immediately. If I relied on freelance developers, well, crap, it's Good Friday. So very least I'm waiting a few days. And then maybe my developer isn't available to drop everything and work on my project on Monday (or Tuesday)... One of the primary advantages of startups is their ability to react and adapt quickly; running a technical startup alone, without a technical background, appears to give up some of that advantage.
-- Not to nitpick, but you are implying the technical team is untrustworthy or incompetent, unless they are co-founders. While I (would like to) think all founders are trustworthy and competent, being a non-founder should not be a contra-indication.
Of course there are other skills - management, marketing, sales, etc. - that would also be helpful. But as a sole founder, while not critical, it seems extremely useful to have some degree of competence in the primary area of work for your business.
In fact, were I planning to start a software company without yet knowing how to code, I think my first step would be to learn enough to build the alpha version (or depending on scope, a core component thereof) myself. Not only is having a specific goal the best way to learn, it would likely result in all kinds of insights into the planned business that just couldn't be gained by farming it out. (In fact, that's exactly what I did with SearchTempest, before eventually hiring a far better developer than myself! :) )
