Obviously, the paper was not intended as such a test, but it works as one nevertheless. The evidence of common ancestry is the pattern in the data, i.e. that they fit a particular tree much better than they fit others. Common ancestry is the only reasonable explanation of that fit. You could consider separate ancestry the null model in that test, and failure to reject the null model would result from data that didn’t distinguish among trees. There are of course thousands of papers that conduct similar tests on these and other taxa. Howe et al. conduct no formal test, but the data are decisive enough that none should be required.
2 Likes