Sorry to raise the dead here, but one thing bears repeating. Since a large percentage (25%) of men are red-green colorblind, it's better to pick two highly different colors. I have to read the original and the modified lines to figure out what's changed -- both "change" colors look nearly the same to me.

Update: I confused the incidence statistic with the inheritance probability. Read a good description of genetics and recessive traits to find out what I was thinking.