Appreciate you taking measurements but IIRC Olights are also known for having cool tints, and from what i’ve read on here meters don’t always read consistently across tints so it may be that while your setup is good for cool white it may not be accurate for warm white. Is that a possibility?
Any chance of making a variable tint version of the Emisar D4?
For variable tint, you would need 2 warm emitters and 2 cool white emitters. (probably SST-20s would work best). They would need to be on separate channels so you would likely need a custom quad star.
But apart from the star, everything else can be done with stock components. Anduril 2 comes with built-in support for variable tint lights so you can use existing firmware.
I have a Sofirn IF25 which uses 2x 3000K SST20 and 2x 6000K SST20 with variable tint. It actually works quite well and is plenty bright enough. The light is just a bit big and lacks Anduril. A D4v2 with the same 4 emitters and variable tint would be perfect I think.
I still say that D18 would deserve more variants (and the corresponding driver upgrades), like this tint mixing idea, or a version with E21A emitters (2 channels of 9A driver, 1A/emitter in total, could be with no tint mixing, or with tint mixing as well).
(Bottom line: Maybe I ask way too much but somewhat bigger (more throwy) TIR optics could be also used in a 10+ emitter flashlight - but that would mean a flashlight head redesign as well).