As the first author of the salmon paper, yes, this was exactly our point. fMRI can be an amazing tool, but if you are going to trust the results you need to have proper statistical corrections along the way.