Perhaps the reference used is not the BBL but the more commonly used daylight locus, which agrees with the BBL below 4000K but becomes greener with higher CCTs. If this is the case, a duv of 0.0021 at 5000K should be below the locus.
This could be the case. In the datasheet it seems that the chromaticity bin is according to the daylight locus but I can’t tell for sure since no BBL is shown in the chart.
Unless you have a lot of red/red-ish object to illuminate and compare side by side high CRI wouldn’t make much of a difference especially comparing lovely creamy R70 5000K to the greenish R9050. In my opinion, especially coming from cool white LEDs, right tint/CCT make much more of a difference than high CRI.
Ahhh good point! I proposed the guess precisely because there is no locus shown in the chart, but now that you’ve said it, the offset between 1C and 2A bins is indicative of some sort of modified locus.
1 Thank