I agree with you that there may be a problem with the asumption of risk adversity and it would be interesting to think further about this. The proposition of the paper is also very interesting, I haven´t thought about it.
But, I don´t think that the study by Boyce, Gordon and Moore contradicts Rawls, actually, I think it is in favor of. Why? Because their conclusion is than people earn utility from their rank in the society. So, if it is true that people loose utility when someone is better, is a consequence of their rank in the society they had in the first place. But, under the veil of ignorance you don´t know your rank in the society, and there is a possibility to be the last of the ranking, and this logic implies that at the end everybody has te same (no ranking).
Then, if you are risk adverse, you want the last to be better of. Rawls doesn´t think that people should like equality by itself necesarilly, its just a consequence of the veil of ignorance and risk adversity, which is consistent with the result of the paper.
Also, it is difficult to imagine the veil of ignorance in a society like ours where social mobility is so limited. If I am privileged, the probability to be the last of the ranking is close to cero, so it is rational for the selfish individual to prefer the things to stay the way they are. Then, the problem is if we live (or we would like to live) under the principle of veil of ignorance or not, but Rawls stands.
Finally, equal treatment of equals is different. If I recall the theory correctly, Rawls talks about liberty and goods, not about utility. There is a clear difference between people having the same utlity and having the same goods. Imagine Alice and Bob having their own utility funcion for a good (carrot), so that for one carrot Alice gets a utility of 1/2 and Bob gets a utlity of 50, then for both to have the same utility Alice receives 100 carrots and Bob just one. Rawls would say that both should receive the same number of carrots. (Please tell me if I am wrong with this statement)