The hacker’s claim is (or should be) that people’s competence should only be judged by their work, not (I’d point out) their worth as people. According to hacker ideology, that’s all that should matter for actually working together in an Internet-based volunteer project.
I’ll agree that personal worth is something different; hacker ideology is silent on personal worth, and I’m not sure hackers would agree on it, the same way they don’t agree on politics. Hackers might claim that in a technical discussion, only technical worth matters, and that‘s a good first approximation. Also, even hacking together requires more than coding skills.
On emotional labor: being in a minority is hard, but nerds are also a minority that must systematically deal with bullying and ostracism. Social justice’s theories to the contrary range from unconvincing to offensive and discriminatory; instead of fighting over what has been worse, we should recognize each other’s pain. And that still doesn’t matter on the limited question of whether, say, your patch should be accepted. How to deal with other interactions is a much more complicated matter; “show me your code” is a way to declare other interactions “out of scope”.
To sum up: hacker ideology is rather successful on evaluating complete code. But hackers miss that this is far from the whole story, and this ideology is unsuitable for a variety of related things. For instance, there is a bias in evaluating coding ability by open source contributions.
Crucially, this ideology can be terrible at mentoring, except for overly thick-skinned people; that’s not only unfair but unhelpful since thick skin is irrelevant to actual coding ability. But expecting programmers to be good teachers in their free time is unrealistic, and arguably even unfair to those programmers; better demand teachers to be good teachers, and education to be universally available. For bonus points, education might even manage to teach hackers how to be more human.
And of course, this ideology seems clearly unsatisfactory for physical meetings, especially conferences open to the public (to pick an easy example).
However, nowadays we do have projects where more human criteria are applied and have proved successful. I think that, for instance, the Clojure and Rust communities seem to offer rather convincing success stories — that doesn’t depend just on CoCs, but on the example set by leaders. The Racket community appears to me (FWIW) to also work rather well without any explicit CoC, but simply due to the attitude of its leaders.