Thanks for being generally complimentary and correcting me on a fine point. What if I defined the red cone as the cone type which, when stimulated alone while the other two types are not stimulated, produce the sensation of light primary red? Would that be accurate even though momentarily neglecting the overlap and the exact graph of sensitivity of the "L" cone's pigment? Ditto for the other two.
Well...I'd probably politely suggest you should have stopped while you were ahead. It doesn't work that way. Hurvich & Jameson found the connection between the trichromatic signals at the retinal layer and the opponent processing signals,
as illustrated here. It is a combination of excitatory and inhibitory signalling from the cones that gives rise to a system that initially seemed wholly incompatible with trichromacy (seriously, red-green, yellow-blue, and black-white opponent process systems, coming from 3 overlapping pigments?)
As for the "cone type that produces a primary red", note that red is an interesting case, seeing as
there is "redness" on both ends of the spectrum, on the violet end as well as the red end. Unless I am mistaken (quite possible, actually), there actually is no single wavelength that is perceived as "unique red" (red without any blue or yellow in it), and we must mix wavelengths in order to produce a "pure" red. (In contrast, there are single wavelengths for unique yellow, green, and blue.)
The overlap is of course essential to seeing the full rainbow of colors. Otherwise, a pure yellow light would look black if it fell between non-overlapping long and medium receptors.
More complex, again. If we only had 2 photopigments, we would be able to match any perceived color with only two lights (at appropriate intensities), and any place where the ratios were about the same (vast areas under our overlapping pigments) would be impossible to discriminate among. As is, our trichromatic vision can be fooled, as your computer or televisions demonstrate quite well, with only three lights. (There is some evidence of quadrachromats, for whom televisions and computer monitors do not look like accurate pictures of the real world.)
Minds -- brains -- a semantic disagreement? When you close your eyes and press them with your fingers and see bright colors, what object is emitting or reflecting those colors? Or if we dream in color, what wavelengths of light are reaching our eyes? The sensation of color is created by the brain. The physics of color is just about wavelengths of light which may or may not correlate to perception, as the castle illusion amply demonstrates.
The doctrine of Specific Nerve Energies (Muller) recognizes that stimulation of any sort of a visual nerve will result in visual sensation; any auditory nerve stimulation in auditory sensation, etc. It matters not what the stimulation is--light, pressure, whatever. When the visual cortex is stimulated (whether by light in the retina, or through brainstem activity, or through classical conditioning), what would you expect it to do besides "see"? When we dream, there is no wavelength reaching our retina; no explanation of vision would suggest it, so it is no surprise and no argument against any physical interpretation.
One more -- what color is my car if illuminated at night by sodium vapor street lamps and it looks white even though it is red in the daylight -- so different in appearance that I can't recognize my own car? Is that white color in the physical world? Or only in my mind?
Color constancy is among the topics vision researchers have studied for decades. Receptive fields (with excitatory/inhibitory center/surround fields) are part of the explanation, but not the whole. My favorite (partial) explanation is simply that what we think happens does not happen; that is, our visual perception is nowhere near as accurate or as sensitive (or as constant) as we believe it is. "Change blindness" experiments are a stunning demonstration of this.
