The tone of the article isn't at all judgemental, Reg.
The LKML is far from a workplace. It's a public forum.
Linus isn't employed, as such, so he's not working and not in a workplace.
And idiocy like this attracts comments like that even in my workplace. Sure, they won't appear on any HR complaint because you make sure they are issued to the right people - like he did here as the response was precisely "fair enough"!
How to cook the article to make Linus the villain again, because he swears a bit. I'm much more concerned about a president who thinks it's appropriate to tweet about nuclear proliferation, tick off China, threaten his staff with the sack etc. than someone saying "This is rubbish, change it".
This BUG() - the article doesn't say it - just kills your machine stone-dead with no way to continue. Sure, it's in debug code, but you DO NOT WANT your machine to just die if it hits a problem. You want at least a way to bail out, an error, and a way to get back to a debugger and test again rather then "Oh well, I have almost no information on what happened, I'll just reboot and hope it never happens again despite the fact I'm supposed to be doing these people a favour and testing their code for them".
And it's picked up by an automated bot, in a commit whose description is "add the option of fortified string.h functions". Things shouldn't bring an entire machine to a grinding half because someone added an option and didn't test it properly (and, reading between the lines, decided not to error-handle, but just to "warn" in the log, then BUG() which kills the machine).
This type of thing might be acceptable in an internal project while testing, it shouldn't be pushed to a kernel where others are trying to get work done (even if that work is debugging that exact problem!).