I don't consider myself experienced enough to give anybody advice on startups, but it seems to me that programmers should pick up non-CS skill-sets. That way, two things may happen.
- Your startup won't be a clone of about 500 other startups.
- You will be solving real-world problems for sectors who don't receive the attention of most programmers.
In other words, follow Octopart's lead.
I know this wasn't a review of your startup, I just don't think this survey idea has any viability anymore.
I generally agree with you, however, from memory, I recall pg saying that entering a crowded market isn't necessarily a move that should be avoided. I think there's plenty of opportunities to beat established web sites by providing better solutions.
In other words, follow Octopart's lead.
I know this wasn't a review of your startup, I just don't think this survey idea has any viability anymore.