So the straight line through the points shows that one stop equals 30cc of neutral density with a 99% correlation.
Pretty cool, that means 30cc = 0.3 log = 1 stop
Then I looked at the package insert for the paper to get an idea of which color was going be the LEAST sensitive. I would choose this one and then dim all the other colors (with ND or RED) to match.
I chose 130cc Yellow as the baseline. I made a contact print of a 21 step wedge under 130cc Yellow to compare to all the others.
Then I used the values in the package insert to guess at some contrast combinations to try. For example I choose 97 Yellow and 17 Magenta and made a test strip with the step wedge. I counted the number of gray segments and got 8. So, on a 21 step wedge each block is 1/2 stop or .15 log. Multiplying .15 x 8 x 100 gives an estimate of the contrast as an "ISO R" value. I got 120 for that combination (about grade 1).
Next I held the processed strip up against the original strip from 130 cc Yellow and shifted them back and forth until the middle grays lined up and the gray values straddled each other. This showed me I was one step of the step wedge off. One step is .15 log which for my enlarger is 15cc of neutral density. So I added 15cc of Red (leave out the Cyan as the paper doesn't see it) and that made the filter pack 112 Yellow and 32 Magenta.
I repeated that process for a number of other values to fill out the chart.
For the two extremes (199 Y, 0 M and 0 Y, 199 M) I just calculated the factor needed to match, again by comparing to the 130 Y test strip and sliding it back and forth until the middle values matched. This gave me 0.5 stops extra exposure needed for 199 Y (factor X 1.4) and 2.5 stops less exposure needed for 199 M (factor 5.6).
I put it all together in a chart and printed it out for reference.