@PurpleBuz
They are based from an Apogee. How are they invalid if corrected? Have you even talked any of the owners/representatives at Apogee on deep red testing? Based on your response, no, no you haven't.
.
I know how to read the apogee data sheet that they publish. its completely cutoff at 660nm .Any valid correction depends on having an accurate spectrograph of the light that you are measuring. otherwise its just a guess estimation based on some other unknown light.. The calibration might be usefull if the apogee actually measured some of the waves over 650 nm in some fashion, but its completely filtered out. the apogee is usefull for comparing similar lights that aren't heavy in deep red and to help a grower optimize their lighting footprint. But you really can't expect anynnmore than that.
If you ever tested PAR levels, you'd realize that the typical PAR output difference between 3000K~4000K isn't quite dramatic and is rather quite similar, with 3000K reaching highest, so comparing 3000K and 4000K isn't exactly out of the question, especially when the CXB is suppose to be much higher in efficiency at that bin. Furthermore, obtaining high-bin 3000K CXB 3070 isn't exactly realistic at this time, so I believe 4000K is a worthy contender given the rest.
If you have ever read and understood the cob datasheets you will note significant efficiency\light level differences between 3K cobs, 4K cobs and 5 K cobs. So how can you not understand that your comparing oranges to lemons and that there are significant diffs between a 3K and a 4K cob ?