This is probably different. I for one don't think there are enough non-binaries out there to warrant a generalised pronoun, let alone its generalised used. I believe even most trans people identify as either "he" or "she".
On the other hand, I like to maintain uncertainty whenever someone's gender is either unknown or undisclosed. When talking about a hypothetical individual for instance, being able to refer to them neutrally is convenient. On the other hand, I hate to use "they" whenever there might be a number ambiguity. In those cases, I'll use "he" or "she", or find another phrasing.
Finally, gender-full pronouns are pretty good at disambiguating roles without having to repeat people's names all the time. Crypto examples with Alice and Bob for instance benefit from saying "he" and "she", so the audience (or student) readily understand whether we are talking about the initiator or the recipient.
Referring to Alice and Bob with gendered pronouns is fine becauae we know their genders. Also, the same logic suggests it's convenient for any third party to be nonbinary.
Yet. Who knows how things will be next century.
I refuse to use singular 'they' for the same reason I don't name variables 'if' or 'printf'.
If a set of people want another personal pronoun then invent one. Don't overload an existing word.
He / she / ke. There you go.
But even a maximally charitable reading of this note - which assumes that Stallman genuinely cares deeply about both gender identity and proper classical grammar, and ignores the long historical use of "they" as a singular pronoun - still doesn't pass the smell test for me. A code of conduct is supposed to articulate your organization's higher ideals and how they are put into practice. A link to your personal blog post about how you think modern language is wrong just has no place here.