Maybe I misunderstand the setting as stated. Would you have to reserve every instance of any potentially problematic word in future usernames and retroactively remove it from community members’ usernames as well?
I know for a fact we have members that use the name ‘testuser’ for instance. I.E this only would work as a solution if we used wildcards to ban usernames en masse, no? That seems like it could easily become overbearing across numerous abbreviations (this particular instance was an ETF ticker in a finance community).