Yes, but why is it new and useful information to know that people might have a zero-sum bias when one is aware one does not have an objective way to decide whether one should act to correct it in any given situation - i.e. should preference be given to the individual or the group, or the group or groups?

Well, in general, being more aware of biases makes one more able to compensate for them if one has some set of goals. Many LWers are utilitarians for example in which case the group, indeed, the largest group, is all that's relevant. In that context, this sort of bias could serously matter.

I find the use of the word "fight" disconcerting. It's a messy word. The way we think is a mechanical process. If there is a right way to think it has the potential to be explained. The word "fight", which carries emotional and irrational connotations, need never be invoked.

This may be a good point. To myself the word fight doesn't have such strong connotations but it may very well do so for lots of people.

Also I disagree with the idea human psychology has evolved to "think the worst" of each other. Can it be proved we have a zero-sum bias? In many situations it would be more helpful to counter positive-sum bias - in Ponzi schemes, for example, not everyone can be a winner.

This seems like a very good point. Foolish optimism of that sort is very common. The Ponzi scheme example is a very good one which seriously undermines the central claim in the essay.