Fence is the expert here for electromagnetic radiation. He would know far more about any of this than I could make a guess at. But I did take some time to study a bit of this and from the double slit experiment to the black box radiation theory, there is little we know about how light actually behaves when it "hits the ground", so to speak. It is reflected, it scatters, it is photons going all over the place and as soon as it is measured or observed it behaves differently!
One of the basis of quantum effects that gives scientists the idea of multiple dimensions is how light behaves as a wave pattern until it is observed, then it reverts from producing a wave pattern to just single photons. Fascinating..
So.. my point, and correct me if I'm wrong, is that without any ability to truly predict how light is going to behave in outside full sun conditions in any environment or upon any object such as this one, elevated very high with a lot of reflected light catching its underside, is very hard if not impossible to put into calculations that would be 100% realistic in a CGI program.
Indeed, if in real life, this is hard to define and predict, how can it possibly be anything but best guess estimate and a refinement of software effects in CGI?