A different interpretation as to why high CRI may be preferred by our visual system is from the predictive coding perspective and how color constancy works in visual perception.
If we take color constancy processes to be our mind trying to predict what the surface reflectance of an object is given the available information, then the wider the spectra of the illumination source, the better the cues our visual system has access to for inferring the surface colors of objects. You could imagine an extreme where you have a single wavelength light source like a ...
My point was that if you were trying to infer 'true' surface reflectance of objects under daylight, while only having luminance info from a single wavelegth light source, you would end up with high prediction error about the properties of those surfaces.
So yes, I was only referring to the case where the observer has a reference with which to compare the objects to. I did not have in mind the case where we would be naive observers having 0 experience with the world around us. My suggestion was motivated by my intuition about our everyday experience, and no... (read more)