Would love some peer review, please explain. Everything there is confirmed by both academic papers, peer research and hobbyist experience. Nothing new or experimental. Not being snarky, seriously, we are always looking to make our lights better, and welcome any and all input and even critique.
Don't use cool whites (if you insist on them since they're 'better' like the FAQ linked above, then use them
sparingly in combination with warm white); don't use dedicated green or red; don't use true UV.
For the 'what's wrong with how we do it'.
Cool white LEDs, especially the ultra cool white Chinese diodes used, are almost completely blue in output, a slight hump in green that trails off before getting into red. This is fine for fluorescent colors, which absorb near exclusively in blue spectra, but for non-fluorescent colors, such as purples, pinks, blues, reds, they require green, amber, and red light, and typically in large amounts. Adding dedicated green, amber, and red diodes can help to a degree, but they are extremely narrow in their spectral output, this is by design. Even a 'measly' 80CRI warm white diode will have a spectral half-width from ~540-640nm, sometimes even farther into red. This is extremely useful, as with a dedicated 660nm LED, its spectral half-width will be from roughly 655-665nm, and if your coral requires light in the 630nm range to really express, then it's too bad, it will not have the capacity to look its best.
As for not using dedicated green/amber/red LEDs, use warm white instead. I already mentioned that a warm white LED covers more spectra than all of those LEDs combined, but in addition, you will not get the individual green, amber, and red color bands in your tank with surface agitation from the diodes being separated. White blends best, and a single diode color (white) will blend better than three diodes. That's just physics.
As for true UV, there is no data that I know of that shows any coloration benefits from it. However, there IS data that shows that sub 400nm (and in my personal testing, up to ~406nm) light fluoresces particulate matter in our tanks that is otherwise invisible. Even tanks with the best of filtration will have these microparticulates, and they are distracting. If you have any data showing pigment production from true UV, I'd like to see it.
Do your fixtures actually use a dedicated heatsink or is the MCPCB just thicker? If it is just the MCPCB, then I would hesitate to say that it is a heatsink.
Where in the hell did you read that Cree uses Epistar diodes? That actually made me laugh out loud! Cree makes all of their diodes and THEY are the OEM for many other LED brands in the world.
That's about all I have time to go over for now. If you have any questions or comments, shoot.