Notes on survey bias.
HN community is international and from [1] it is around 3.5M monthly unique users. A representative sample would need to be stratified since it is unlikely that — for example — the proportion of female readers will be the same across countries.
The poll would need to run for a long time since the daily and monthly unique figures are different by order of magnitude.
Further, there are many times more lurkers than posters [1], and only a fraction of posters will answer this poll (I will not, for example). I would expect something like a >99% non-response, implying strong self selection amongst responders.
The upshot will be that inference from this poll to the HN population will likely be biased.