To be fair, Linus did change. That charge about not valuing version control and systematic release practices? He built git in 2005. Kernel release practices are now a well-oiled machine. And overall the kernel has likely got more modular, not less, over the years (although I have not had a look in a while, but that was the trend a few years ago).
ESR might have been wrong on the specific item at hand (I have not read the whole thread, and he, ESR was wrong on a lot of things over the years) but some of his arguments are definitely right - at least seen from a "gifted something" who flopped in uni ;)
But was it due to ESR's ramblings or due to patches overflowing his inbox capacity?
Personally, for years I had been completely ignoring everybody who framed something as "a better alternative to talent" (wtf?) or said the magic words "you are smarter than me". They correlated too well with "stupid ideas I knew (or quickly checked) to be wrong".
Nowadays I'm trying to fish for useful information even in messages which push my bullshit-o-meter out of scale, but it takes work.
If you want to maximize your chance of being treated seriously, stick to constructive arguments about the problem at hand, not some generic unfalsifiable wisdom.
The reason for that way of phrasing is because it appeals to ego. If you want someone to do something, particularly someone with a known ego, then this is an effective way to get them to listen to you and not dismiss you.
Of course, there's a fine line between stroking and patronizing, though.
It works until you get to those with really big egos who take pride in dismantling such tricks and gaming others by pretending to fall for their games. These need higher levels of gaming.
But in the end, if you truly think that something is a good idea for everyone involved, just explain why.
Even before git, he started using BitKeeper in 2002. I guess Linux realized he couldn't continue without some more structure. Even if he didn't need it, other developers do.
Well he did in part because at one point Alan Cox's branch of the 2.4 kernel was threatening to become the de-facto kernel. This because Torvalds had built up a massive patch backlog.
ESR might have been wrong on the specific item at hand (I have not read the whole thread, and he, ESR was wrong on a lot of things over the years) but some of his arguments are definitely right - at least seen from a "gifted something" who flopped in uni ;)