A 20 year old is also much more likely to not worry about death, and be unable to spare a thousand bucks a year or so. As well, modern 20 year olds come from an era where cryonics is a joke they see on TV (Futurama), and not a real possibility like it was for people at the start in the '60s or '70s.

If your age inference is right, shouldn't we see a lot of young people in cryonics? But recall that one of Eliezer's cryonics was about a cryonics conference aimed at recruiting young people; not the sort of thing you do if you're reaching them very well... This also lines up nicely with my previous post about the increasing cost of cryonics due to ending grandfathering: it was previously supportable because cryonics was growing, but now...?

gives you more categories than (standard) gender, subdividing by decade say gives you 5+ categories in the adult population not 2(ish).

It also means your inferences are less reliable because your total n is being split over 5+ groups and not just 2.