I think the demographic for lemmy (and lemmy.ml in particular) is so niche and so biased towards privacy that even with a bigger sample, it’s next to impossible to get really good data. Probably good enough to say that it’s very likely that there’s a significantly skewed gender distribution, though. The user polls that were done on my instance showed similar results IIRC.

How could that possibly lead to a bias? /s

I played around with cubic, if I recall correctly it worked pretty well and wasn’t super unintuitive. It’s based on Ubuntu.