If I'm missing something, let me know, but it seems to me that GitHub did about all they could with this investigation (respected outside investigator, seem to be releasing as much as they are able to, although if they stop releasing new information/new stuff comes out through other channels, my opinion will change). They've admitted that TPW's conduct was unacceptable and are trying to correct that from happening again. The allegations about sexual harassment and locking Horvath out of code seem right now don't seem to have any merit beyond Horvath's allegations. Also, Horvath seems to have accepted the outcome, simply noting that she disagrees with the findings, but not raising issues with how they were arrived at or trying to impeach the investigator's credibility. Her belief that none of this would have happened were she male is an opinion she is entitled to, but is basically impossible to prove at this point. It's unfortunate, but, beyond GitHub having a time machine to go back and try to fix this as it happens, I'm not sure what else anyone expects them to do. My hope: GitHub continues to release what they can, they improve all their employee mediation processes, everyone agrees that they made big improvement to their internal culture, and that eventually, this whole ordeal is relegated to the annals of a Wikipedia subsection.
You said it perfectly. However, I would add that it seems to me that at first we give more credence to the accuser than the accused. The burden of proof in this situation is always with the accuser and the fact that she assumed people would believe her, and people have, is not something I respect. And being a male has no effect on my opinion. I have pissed off my friends and especially my family, father in particular, because I've always given the benefit of the doubt to the accused in similar situations. My father found it particularly annoying, rather extremely frustrating, that I wouldn't automatically support and accept anything he would say at face value against a business partner or relative or anyone. And my behavior is the same with friends and coworkers, much to their annoyance.
Back to point of Julie's complaints --- we're adults here and if there's a pattern of behavior from a coworker, then documentation is your best friend. In fact, that should be an automatic rule: any communication with colleagues in the workplace should be documented, either via email or chat, and kept short and to the point otherwise. If there exists a friendship between coworkers and there is fallout due to conflict, assuming there is no evidence, the accuser/accused should just let it go because that's what friends do. You can't demand professionalism when the principle hasn't been observed in the first place.
Being a man, woman, father, mother, doesn't absolve one's responsibility from presenting evidence for the truth, even if they have a history of being honest. More importantly, our attitudes towards the accused should initially be of giving benefit of the doubt.
Consider how what the cost is to Horvath here, how she can say whatever she wants, launching a huge firestorm, and oh, 1/3 accusations is provable. (And one, about the guy removing her code, is actually pretty much disproven.)
Continuing to call the workplace "toxic" is objectively and clearly an overstatement as well.
I generally believe that, when I have no personal knowledge of either party, or the validity of the allegations, innocent until proven guilty is a somewhat healthy mindset, although the burden of proof for "guilty" should likely be somewhat substantially less than in a court of law
Have I missed anything?