This isn't quite safe harbor. It's not like the app was removed for one user posting one bad content. If what the poster above said is true, it's closer to if an app had a user who regularly broke the rules, and the app refused to ban said person.
I disagree about your comparison. This app can connect to arbitrary ___domain names. This is getting blocked because you are not filtering the list of domains a user can connect to proactively.
That's wild & I can think of zero precedent for it.
I'm not sure what you mean by justification. I think I simply lay out some context and a set of conflicting perspectives.
That said, if you don't want Chrome and Firefox to be content aware, then you should argue that safe browsing should be eliminated from Firefox and Chrome. That is a self consistent position, but it may not be consistent with e.g. avoiding dramatic growth in botnets, ransomware, organized crime etc.
Actual safe browsing comes from content-unaware tools like NoScript. And yes, I did spend half a hour going through about:config and neutering everything related to 'Safe' Browsing(R)TM(C)LLC.
This is grossly unacceptable. Apps need some safe harbor too. Apps can not be responsible for every possible use of the app.