There's another explanation too, which is that if you pay attention as you age, you find that nicely wrapping up all your engineering points in a polite wrapper can very easily put you into the unenviable position where half your audience doesn't correctly penetrate the circumlocutions to understand what you were trying to say, while the other half goes right on ahead and gets offended anyhow, because what they are offended by is that you are disagreeing with them at all, no matter how much padding you put on that.

I'm probably still too verbose (ha... "probably"... I flatter myself), but I went through a phase in my late teens and early twenties where I qualified and wrapped in polite circumlocutions everything I wrote. It makes in incomprehensible, honestly, and it doesn't work anyhow. I still got flamed. (Good ol' Usenet.)

There's a time and a place for politeness. That time and place is most times and most places. There's a time and a place for directness, bluntness, and even outright impoliteness; I'd submit "Intel is just going to leave a critical vulnerability in basically every CPU in the world, and throw up an obfuscating cover of flags and workarounds and PR to try to pretend otherwise" is a reasonable time for those things.

(Personally, I don't expect miracles. Intel ought to come out and say "Look, this is going to take some years. Here's some workarounds in the meantime, and here's our plan going forward." I think everyone who understands the situation understands that a true fix is very difficult. But if it looks like they're going to try to skate by on some weak workarounds and a heavy dollop of PR, the industry as a whole really needs to pushback, not just Linus.)