I'm guessing the article refers to a form of "cowboy coding" or similar things promoted along with the myth of the 10x programmer, where there are attempts to apply that practice (of siloing developers and having them only write code to satisfy their needs) into the organizational environment. And I'd agree with the article there, in my experience that's awful and leads to everyone being upset. The hermit programmer is no different from a rogue programmer. It's a great way to ensure that key pieces of information are missed or glossed over.
It all depends on what your perspective is. This article seems to speak to web devs, and I agree with their premise for the most part. But it doesn't work for single indie game developers. Hermit programming is very effective in that space (insert Stardew Valley reference here).
So I come back to... why bother even writing this? It's no secret that poor "team players" are a nightmare to work with.
I'd be careful with that, I don't see why game development would be any different. You still have to focus on customer satisfaction. Single indie game developers succeeding without doing that seems to be an exception, not the rule. For every Stardew Valley, it feels like I've read a dozen postmortems from people who went off developing something for years only to find out that nobody wanted it and there was no audience for it.
