IS “SOME BETTER THAN NONE” OR DO ALLELE FREQUENCIES CONTAIN PHYLOGENETICALLY USEFUL INFORMATION?