As valuable as I figure he is to the project, I don't think it depends on de Raadt per se.
Basically he has a bunch of people around him, whom he trusts-- and he hates grooming favourites (which maybe isn't that typical for people as opinionated and meticulous as he seems, though it's probably good news in this context). Or in other words, he hasn't made it Theo-dependent. I guess he's smarter than that, and good for him really.
No citation sorry, but I watched an interview with him from a few years ago, and this is the impression I got from the interview.
On of the things I love about OpenBSD (seems relevant to this question) is it's probably the most forkable general-purpose OS there is. I read that Linus doesn't even know how his own kernel works anymore. That's too much complexity (IMO).
Maybe OpenBSD isn't designed as completely "general-purpose" but it certainly is by the modest standards that matter to me. I'm using it that way and I'm not the first to do so. It would be even more general purpose if it supported more architectures.
