The six proposed assessment metrics are compelling reading; I wonder whether there's data out there to support/reject the effectiveness of each measure in widely-used software?