Any system can be abused. If they were turning arXiv into what amounts to a Facebook comment section on a late Saturday night, it can spread rapidly as we’ve learned over the last few years.
There is of course the _potential_ for stifling scientific ideas. I don’t know if there was in this case. arXiv claims it did not stifle. The person banned for six months and some friends believe it has or could. Certainly the person banned for six months is unhappy.
In my opinion, a better solution might be or have been to put any questionable papers such as unprofessional inflammatory personal political debates into their own section; a “lower tier” still available, but in a place where it’s very clear that this is distinct from the main area.
But it can’t be arbitrary; this case can be used to construct written guidelines of professionalism that don’t affect any other papers; something about this must be unique to this situation that entered the arXiv space that hadn’t been there before, otherwise it is simply “removal at whim” and whim would be bad.