HN is also a fraction of the size of Reddit, so there is less value in attempting to do so.

And it has zero incentive to use its data other than to sell joining YC to entrepreneurs.

Psychological techniques tend to work no matter whether we are aware of them or not. Daniel Kahneman's "Thinking, fast and slow" explains why this happens, with lots of examples. It's a great read.

I'd argue that's hard to measure.

Can you convince a subset of doctors to go anti-vaxx as easily as you can a randomly sampled subset of laymen?

I suspect much of the readership of HN isn't only aware of dark patterns, but actively primed to recoil from them.