Right now people with an interest in making the world a better place have sort of two loose sets of social rules they can draw from. There's what I think of as "liberal norms of discourse" which include fairly free speech, expectations of polite disagreement across a wide range, and some variety of bad-faith good-faith. These work pretty smoothly across a wide range of views and levels of mutual trust, but it's fairly easy to recreate existing systems of oppression within them; people with extra physical and psychic resources tend to "win" these kinds of debates, and lots of harm can be done to vulnerable people below the threshold of concern for liberal discourse.
The alternative is the loose collection of social justice standards, which generally exist as a response to a way in which liberal discourse fails marginalized participants. Censure of microaggressions and focus on inclusive language is supposed to lower the threshold of concern and prevent a lot of harm. An absence of tone policing is supposed to allow those who are very personally affected by an issue to talk about it without dismissal. Elimination of bad-faith good-faith conserves energy in the face of a surfeit of what are mostly bad-faith interlocutors. All of this works pretty well in a high-trust environment and is often a much better experience for people who are shortchanged under liberal discourse. But the unboundedness of most of these concepts produces an environment where anyone is guilty of some level of transgression, and without the trust and proportionality of both parties knowing each other personally even minor disagreements between reasonable people can escalate to absurdity.
This leaves community creators with a kind of sophie's choice; do you choose the effective standards of liberal discourse and just live with the fact that some people won't really be able to participate equally? Or do you try to work with the ideals of social justice and hope your community stays high-trust enough to keep things reasonable?
(In response to +Andreas Schou
's post on this article: https://plus.google.com/u/0/+AndreasSchou/posts/T1SckNZJzmr