GST is still a useful measure of genetic differentiation — a comment on Jost's D