Hi Matt - I enjoyed your paper a lot. Two quick comments. MP-EST can take alignments with missing taxa and you don't need the same outgroup for all alignments, so perhaps you could include more genes than you did in the MP-EST analysis. Second, there is a cool way to conduct topology tests similar to CONSEL in a coalescent framework using MP-EST. The ability of MP-EST to calculate a likelihood is one advantage over ASTRAL, which doesn't have a likelihood model behind it. This might be interesting to try, please get in touch if you are interested. Again, great paper!