I’m building four Brinyte B158 aspheric throwers currently, and it has been a good way to compare the results from various different modifications.
One thing that really surprised me was the intensity of the dedomed XM-L2 U2. The dedomed XP-G2 S4 2B (the new, not-as-good one) gave 238kcd. The dedomed XM-L2 U2 did 220kcd, with a much larger die image! And the XM-L2 was actually on a very thin copper board, not the thick Noctigon the XPG2 S4 was using.
The bottom left is the dedomed XPG2 S4 2B (238kcd), bottom right is the dedomed XM-L2 U2 (220kcd). Top is an XP-L Hi V3 3C. All three have MTNDD drivers and everything else is the same between the three lights also.
Sorry for the bad pic, I took it really rushed because I was worried about the heat sinking on the XM-L2 (see my recent threads for stories of fried LEDs).
Now I need to build an integrating sphere so I can measure the output difference in flood mode. (in a more efficient manner than my current ceiling-bounce-integrating-bathroom setup)