If that's the test reading, these xpg2 with AA are running brighter than the AA XML2 mule,
xpg2 lumens: low -6, med- 17, High -110
XML2 lumens: low -2, med -10, High - 85
This is why I would prefer not providing any numbers at all. The proper way to measure lumens from a light is to have a sphere that is 2-3' in diameter and have the light inside of the sphere and have the sphere and spectrometer calibrated as a complete unit. My sphere is 6" in ID and I introduce the lights to it via a port on the side. the wide flood of the Mule likely does a much poorer job of getting the light into the sphere compared to the haiku reflector and the smaller die with tighter collimation with the XP-G could give it the edge over the larger XM-L2 die. I had to change to a smaller fiber optic cable that connects my sphere to the spectrometer when I attempted to get some quantitative measures from the Aqua lights so relative relative measures between lights taken before this change compared to measures taken after this change is that much more suspect.