Since when is sharing a platform with people whose political views differ from yours equal to allowing targeted harassment? I don't think many would disagree that someone who threatens real life violence on someone should result in a banning from that platform. But an automatic ban based on political affiliation seems undemocratic.

The original article doesn't mention anything like "an automatic ban based on political affiliation". The de-verification of white supremacists argument hinges around both verification as endorsement, and the possibility that some political beliefs (white supremacy) are in themselves a threat of violence.

Meh, these threats exist in the real world as well. You can't silence everybody who might possibly be a threat at some point.

Those who actually make the threats should be dealt with, but last time I checked it's "innocent until proven guilty" not "guilty until proven liberal".

Yeah, it's pretty fucked up when an internet lynch mob gets you fired because you said "dongle" or something similarly bad.