Neurons (and photoreceptors are neurons) tend to code events as a change in firing rate. All neurons have a quiescent firing rate.
If you tickle the retina with flickering light you get subjective colors
http://www.google.com/search?sourceid=navclient&ie=UTF-8&rls=GGLG,GGLG:2005-31,GGLG:en&q=benham+colors
I'm thinking the flicker rate that produces the experience of color must be related to the firing rate that would produce the "true" experience of color. I did some work on this in college, but it's an incredibly difficult field.
Based on what I know, the trick is to stimulate the cones in such a way that the brain perceives a given color. There are three types of cones, each with a different spectral response. I can easily imagine that, as the disk is spinning, the constrast line between the white and black would move across the retina as a front. Since the cones (receptors) are activating at slightly different times, because no two receptors are in the exact same location, then the brain perceives this phase difference as a different ratio between different receptors and color perception is generated. It is a very interesting illusion.
I've often wondered what would happen if one simulated one color's code with a flickering light of a different color...