Experts in one field are amateurs in all the rest of them. So even a community of just experts in some field would have a much larger number of amateurs than experts across all fields.
That's true but I'm not constricting "expert" to just domain expertise. An expert in <whatever_technology> who has worked hard and invested time into acquiring that expertise is often also an "expert" at not being a nuisance in the fields that he's an amateur. He's able to ask quality questions (instead of inviting snarky lmgtfy.com). He's an expert at being a non-expert.
Even if Paul Graham doesn't know Rust, I think it's safe to assume that any questions he'd write in an online community would not be ridiculous (is Rust slower than Bash?). Any amateurish observation he'd have, at minimum, would have some intelligent thought behind it.
HN is mostly very civilized. (Granted, some of that is due to moderators.) So the blog's bullet point about "trust" devalued because of one-time transactions does not seem relevant. It's relevant for reddit/atheism but not HN.
To me, the problem with HN, SO, etc is the voice of experts getting drowned out by amateurs. The growth rate of new amateurs joining will always outpace new experts and each year the signal-to-noise gets worse. I haven't seen a clever self-governing mechanism that addresses this social dynamic. The upvote/downvote/karma history is not enough to solve it.