If you were going to normalize the curves, why not do them so that they are comparable to the ones last year, that is to reflect a 1W 1M equivalent as opposed to something one has to re-calculated for each curve.
Also, if you check those boxes here on the comparison page, and account for the 28.3V drive into a 4 Ohm load, these are still too high in comparison to last year and JBL's published data.