Well.... as I recall - the readings were taken 1-2 inches from LED array (maybe 1" from nose or 2" from cheek). The Omnilux read higher than Lumiere mostly because there were not any divider-bars between the clusters of LEDs (I think).
As with tanning beds for UV readings... illumination geometry is important (canopy curvature, reflectance, etc), and of course amount of LEDs plus how closely packed they are. Also input wattage and yes individual LED output in the 633 ±xx nm bandwidth SPD narrow bell curve vs meter responsivitity:
Got all that? Does Underhill lol?
I have NO foggy idea what the TA or ProSun red LED units put out.