You guys continue to underwhelm me; to compare the classics to the moderns is quite simple.
First and foremost, you've got to run a regression analysis of the modern criteria vs. the classic criteria
Upon gathering the results of this analysis you should have ample data to derive to correlation coefficient between the two.
By applying the correlation coefficient to the disparate criteria you can then geo-metrically link the two to come to an approximation of the results.
However, approximates won't cut it when it comes to golf course architecture analysis. Therefore, you've got diagram the results into chart and see where the slopes begin to demonstrate convexity. This will begin to narrow the approximation down to very managable and discernable levels.
By taking this measure of convexity and double checking to ensure the results demonstrate no heteroscedasticity, then you have comparable data within a normal distribution pattern.
And voila, you have your apple to apples comparison for Golfweek's modern and classical lists.
Easy as pi!