a) obviously weight trusted mods before rando community members who might just hit random buttons to access the server.

b) but I feel like that would give the mods some guidance?

Literally just add all the mouse movement or whatever goes into the captcha, but also collect data that suggests whether or not a given post is neo-nazi shit or w/e.

Seems like a win-win? More helpfully flag community violating content while ALSO verifying non-bots.

Just-sayin’