OK so I've just been back down and put the GCCP paste back on the trunk lid so that the whole car matches (P21s was too light a reflection). While I was there I decided to play a little with the lighting, and see if there was any way that I could repeatably identify this uneven finish, especially with artificial light... And I succeeded!
Normally after finishing waxing, we (detailers) look at the car in the sun, or using a point light source like a Brinkmann Xenon / LED flashlight. I too have tried this, but as I said this effect is not really noticeable in bright sun-light, so that doesn't help at all.
Highly polished black surfaces are almost like a mirror, so instead of a point source of light, I tried a highly diffused, but bright source... The Brinkmann Xenon pointed at the ceiling in the Garage.
Instantly the whole surface of the paint changed appearance, and it was like I was now able to look at what was on the paint itself. Dust particles were clearly visible against a sea of dark grey colour. Because I was only looking at the surface, I couldn't see the metallic grains under the clear coat, but what I could clearly see were the different colours of the coating.
This time it was more obvious that they followed the outline of the initial application of the polish, and like I said before, it was more noticeable where the wax had gone on thick, compared to thin.
I would like to challenge anyone else to test GCCP using this approach, and see if they are EVER able to get a uniform coating.
Maybe this is the same for all LSPs? Maybe I'm only now noticing this problem? Maybe this is the reason that people talk about layering being important (though layering doesn't seem to help with this problem to be honest).