I'm hard-pressed to discredit someone who got Digg the attention it got, then maintained and built a personal following that he unquestionably influences, and THEN spearheaded a product that you may not love but did get 150k users and was damn beautiful to boot.
I'm not a fanboy, but Kevin Rose is on a short list of people who can really fuel a product and it doesn't require technical knowledge or even 'having the idea' for that to be valuable.
You don't need both legs to run a marathon. But it's so important to have them both. That it's no coincidence that every great runner has them both.
To me, there's actually a difference between knowing how to code and knowing the technology. You can know how things work and what's technically possible without really knowing how to code it.
But this goes both ways too: developers don't necessarily have a good product sense, and without knowing about interaction and design patterns for example, or know what the current product landscape looks like, they would also be limited in their implementations. My comment above was mostly to contradict comments like the parent, that are fairly common: hackers can do it all, knowing how to code is the most important thing. Coding is not important. Knowing what can be coded is.
The "idea guy" is the guy who comes to a developer and start with "I've got this great idea for an app…", thinks it's genius but never dives down to all the details of said-app.
A product person might or might not know how to code but s/he will(/should) be able to articulate exactly what the product should do or not, how things should work together, etc. It's common for developers to think they don't need anybody to make a good product but a great product person will make a difference.
It doesn't matter if they understand how to code or not.
Indeed. That's not their role. Understanding how to code does not a good product make. There are tons of examples of that.
However, you have a point about the industry giving a pass to people with previous successful project(s). But probably with good reason: overall there aren't that many very successful projects, so if you had one, it's still much better than a lot of people.