You’d be surprised how harassers can use seemingly innocuous actions to harass someone.
Eg. Here the example of a vindictive ex or bully was brought up. Now, on the surface someone like that isn’t doing anything technically wrong by submitting IDs…but they are subtly telling their victim that they are still present, can find them (assuming the observation’s location isn’t obscured), and possibly using dissenting IDs to imply that their victim’s opinions are invalid.
To your point re: banning outright, if there is no paper trail (restraining orders, etc) then how does iNat know which party to suspend in such situations? In that case, blocking makes sense.
Honestly, all that was all stuff I hadn’t even considered until that thread mentioned bullys and exes.
[addendum: fixed link]