He's using internet speak. Since it can't be communicated as such in any other way, it's likely an exception, although to someone who hasn't been read in it probably reads funny.
He's not wrong, but he's not taking into consideration the fact that it's a feedback loop. It's the same perspective as chiding the homeless for not pulling themselves up by their bootstraps. They are in too much mental poverty to do it.
It's a masochistic method of programming where you insert a randomly generated layer of abstraction between the hardware and the programmer, and tell the programmer to make sure to add enough hints to his code to make sure that it runs quickly.
That's a stretch. If the only dimension you care about is whether or not the language spec is guaranteed to be invariant, sure, but that's essentially nitpicking. One more layer of abstraction gets you an asm->asm transpiler that never changes.
It doesn't matter that it's not cryptographically signed and not using https, etc. If the paper contains details about the history of bitcoin that can then be proved to be true, that is well on the way to validation. i.e., details such as "timechain" preceding "blockchain."
In your own words, you are asking for 'science'. And if you click the link through to the SEP, you'll see that it's well documented that this is, in the words of Stephen Wolfram, A New Kind of Science.