Just so I understand: You're always willing to trade intelligence and experience for someone who produces more, for essential key components of your startup. So you would hire a less experienced person who writes more code over a very experienced slow developer, do I have that right?

All that does is trade the short term risk of not having as much developed with the long term risk of having a broken system. It means you're potentially betting your business on someone who isn't the best, just because they seem more productive.

In summary, you risk making a Friendster instead of a Facebook.

Let me know if I misunderstood your point.

I hire on both: smart people who get it done. And getting it done is relative. Unless you're in consulting, velocity is almost never the issue compared to correctness, so a smarter, slower person is sometimes the right choice. A member of my team took a long time to come up to speed, and is now kicking ass and making the right decisions to keep things on track.