We haven’t run that exact experiment, but in the 2-party setting (1 strategic, 1 honest), there is a known threshold of hash power below which selfish mining is no longer profitable, which can be computed (roughly 0.25, but depends on some parameter settings) (Sapirshtein, Sompolinsky, and Zohar, https://arxiv.org/pdf/1507.06183.pdf). For this reason, my guess is that in the setting you are describing, it would be similar to 2 strategic parties and 1 honest party with just a little bit of hash power. And that actually is an experiment we ran (https://arxiv.org/pdf/1912.01798.pdf, Fig. 7). Here we see that as the two strategic parties get closer and closer to 50-50 hash power, their advantage from selfish mining vanishes. And actually, they seem to not be stealing from each other, but from the 3rd honest party, while converging to an honest strategy only when the hash power is actually 50-50. So I believe SM continues to be profitable for 2 strategic players, but our experiments suggest this may not be the case for 3 players.
3 Likes