On the desirability of models for inferring genome phylogenies.