Comment on W. Erwin Diewert’s, “Index Number Theory Using Differences Rather Than Ratios”