I don't necessarily think you've misidentified the culture. I
disagree that it's sick. I think it's not only exactly what we want to have, but a
huge improvement over what we started with in 2013-2014, when new question volume was peaking, curators and experts were getting increasingly frustrated, and the volume of discussion on meta exploded. (I've noticed that whenever I need to refer people to Q&A on meta that's our highest-quality
meta content, to explain what Stack Overflow
is, a large fraction of it is from 2014.)
In particular: we have always had what could broadly be called a code of conduct; it's become more refined and more like official codes of conduct over the years, for better or worse. But overall, over time, we've become much better at removing actually abusive, profane etc. comments, and editing off-handed details in questions to avoid giving needless offense. (By the way: a quite large fraction of curse words and insults come from new users who are upset at the realization that questions are subject to quality standards, or who take downvotes personally when we intend it purely as content rating.)
When I say that I don't understand, it's because you describe "in-group out-group aggression and defensiveness" and I don't see it that way. I'm not trying to protect other meta regulars. I'm trying to help people integrate by explaining to them how we want them to approach the site instead.
But it's impossible to do that without first informing people that their current approach is wrong, and trying to explain patiently why it's wrong.
> it's what happens when people view themselves as lone defenders of something they care about.
Because we actually, objectively are.
And what's wrong with that?
Why shouldn't we be able to have this thing?
And why should it be considered an invalid thing when e.g. Wikipedia is not?
If 29 million people want to use the "anyone can edit" property of Wikipedia to edit https://en.wikipedia.org/wiki/Dog and ask whether Rover's condition is serious enough to require veterinary attention, does that invalidate Wikipedia's model?
> that begins to actively degrade the plebian outgroup that they started out serving.
Stack Overflow started in a closed beta and was marketed from the start as being for people with a certain level of cluefulness. We had to argue among ourselves to get everyone to accept that a) easy questions are not only fine, but often the most valuable and b) the thing that experts tend to hate about beginner questions is not the fact that they're beginner-level; it's literally every other consequence of a beginner asking them.
And acceptance of that is still not complete; sometimes long-standing members get yelled at on meta for trying to close good, easy questions because they're easy. And they, too, are acting against consensus, and against Stack Overflow's vision. (They're just, you know, nowhere near as troublesome overall as the long-standing members who don't care about policy and just try to answer as many questions as they can figure out an answer to.)
Stack Overflow was never intended to provide the kind of "service" that most newcomers (including newly arrived experts hoping to answer questions) expect. It was instead intended to show people that there's another way, that's fundamentally different from the traditional forum experience.