You don't have to agree with a criteria to be able to apply it objectively; it just has to be something that can be evaluated by anyone to come up with the same answer. You can't objectively evaluate whether art looks good; you can only get subjective answers and attempt to find consensus. You can objectively determine if a word is a noun, verb, adjective, etc (given context, as some words can fall in several categories). And given context, you can objectively determine if a word is a slur. (Personally, I avoid both slurs and insults, but github only bans the former, not the latter, and that seems like a reasonable position to take.)

So, to be clear: there's an objective criteria for "what is a slur"; there's no objective criteria for "what is acceptable" or "what is offensive". But in any case, a hosting company is hardly required to provide an objective criteria for what they accept; it's nice if the rules they supply can be objectively evaluated, but that doesn't mean their choice of which rules to apply needs to be objective.

In any case, I have no further interest in arguing with you; I doubt you're going to stop being outraged that someone won't give you a free platform for hatefulness.

I already have a free platform for hatefulness, it's called America.