When working on my own led head, I reached the conclusion that "normal" (ca grade 2) contrast is definitely not a 1:1 blue:green ratio and that the green emulsion of VC paper is a few stops slower than the green (which makes sense, of course). Equal amounts of blue and green resulted in roughly grade 4 in my experiments. Hence, I think your best bet is just to plot some curves for the different contrast settings of your head, as there will be some non-linearity in the system here and there.