I think you're right about the HN-bait stories -- in fact, I have a feeling that a number of people tend to upvote sensational stories (actually, just the titles) without even reading them, which makes the problem even worse.
I assume other people do the same. It's not a good thing to do.
I think the site would be much more useful if it were just Ask/Tell/Show HN stories with some kind of meta filtering on top. How do I find a cofounder? What's the best way to measure traction? How do all these startups get so much media attention? Is there some way to navigate from those questions to all the discussions about it that we've had on HN? If so, and if you could do that with a 100 startup-related issues, it'd be a hell of a more useful site than all the gossip and drama that we end up with.
It occurs to me that any story that takes a long time to consume -- say a long TED talk -- has a built-in disadvantage to a story that can be quickly consumed. That's because of the way the decay rate works.
Ideal HN-bait story? A long-ish story with several technical-looking drawings, charts, and graphs, all around some flame-bait brand/fanboy thing like Apple or Google, loaded with personal drama that involves some famous tech person. If it has some technical googledegook or obscure jargon, all the better.
I don't mean that as a criticism of HN, just trying to flesh out what commonalities I'm noticing. Seems like that would be a first step to critiquing site. For all I know, this is exactly the kind of story pg wants to see. Beats me.
EDIT: This idea of trying to come up with the silliest hn article title in order to demonstrate community trends was so interesting to me that I followed up with a HN Contest http://news.ycombinator.com/item?id=2048844