Earlier quoted context omitted.
It got crossposted on various places that have little/nothing to do with programming, so hopefully most of the nasty stuff was just from random dickheads rather than the python community.
Bingo. The meanest reactions have been people with a particular axe to grind rather than Python community members.
Public shaming is a very powerful thing, and not inherently good or evil. Some people need to be shamed for what they've done, and others need the threat of shaming to keep them in line. But shaming's power makes it dangerous, and not to be used without very careful, independent verification that it's warranted: that wrongdoing was indeed committed, that it was committed by the target, and that shaming is both appropriate and necessary as a consequence of that wrongdoing.
The "independent" part of that is important. Even if we discount the possibility of malice, there are too many ways for things to go wrong. Definitions can be unreasonable, identities can be mistaken, memories can be faulty, evaluations can be based on incomplete information, and so on. The involved parties are, in all but the most egregious cases, not qualified to make those decisions unilaterally: they're too close to the incident. In all but the most egregious cases, the involved parties are too close to the incident to be qualified to make these decisions unilaterally. Even in the most egregious cases, the very things that make one side qualified to make some of those decisions also disqualify that side from making the others, which is why we have courts in the first place.
I think PyCon's actions here are quite appropriate, and a marked improvement from the earlier code of conduct. There's room to improve even further -there should be some basic outline of a process for reviewing and investigating incidents- but this is still a step in the right direction.