I'd say there is logic in not bothering to make the change, it takes effort and brings no value. My experience has been that gendered pronouns are often used in documentation, though usually they use a mix of the two genders.
That said there isn't logic in attempting to revert the commit when another committer merged it (unless the rule was that documentation would use a gendered pronoun). Though I also don't think the tone and content of Joyents post is appropriate either.
So unfortunately this means there really aren't any open source alternatives.
Unfortunately the distributed model has fundamental privacy problems. One needs complete trust in all server nodes as they can do nearly anything with a user's data after they have access. e.g. a user can revoke permission but that doesn't prevent a networked server from having cached & continuing to display it. Or potentially a rogue server which makes everything they have permission to see publicly available.
"If I do not want others to quote me, I do not speak." -- Phil Wayne