I really had a good idea tonight, having a whole evening for flashlight hobbying (=scarce) I wanted to do an output test tonight on the 7A3 80CRI XP-L U5 bin led from intl-outdoor, and then again dedomed because I wanted that led dedomed anyway for a triple together with two 2A XP-L leds. It would give more data on the XP-L in general, and I could check the strange result I found before with the 2A emitter that dedoming did not seem to result in lower output for the XP-L.
Then the luxmeter in my integrating sphere gave reading zero. Checked everything and finally disassembled the whole sphere (the sensor is permanently attached to the inside of the sphere) to find that a wire had broken right at the solder joint on the sensor board (I had fiddled too much with the sensor wire over time, so it broke). Soldered it again, also made a new aperture before the sensor because everything was open anyway (the old aperture was attached in a kind of makeshift way, ok working but the new one fits better), assembled the whole thing, re-did all the necessary calibrations (the values did not change much, a tiny bit) and I was living two hours later 
.
Started the test with the XP-L 7A3 80CRI, all went well, I was going to stop at 8.5A and then dedome it. Then between 8 and 8.5A:
The bond wires blew early. I have heard before of leds that blow at lower currents than I had measured, so apparently there's variation in that, but I had not have the pleasure myself yet . As you can see below it is indeed the bond wires, where a clear wire should be is two blobs of silicone:
So now it is bed time, repaired the sphere, lost a very nice emitter and gained nothing.
Well, there is the test series of the XP-L 7A3 80CRI up to 8A. Summarised: 1) the output of thw 7A3 80CRI U5 is 20% lower than the 2A V6 led (as quite expected), 2) the maximum output is at 7.5A, as with the 2A tint XP-L, and 3) the voltage of this specific led is 0.2V lower than the XP-L 2A led that I used in my first XP-L test. That is a large difference, this 7A3 led comes very close to the voltage of a normal XM-L2, while the 2A that I measured earlier had a surprising high voltage. I guess what I see is voltage variation between leds (does Cree do voltage binning?).
I hope that the next test session goes better...
May that led be an offering for the Lumen gods and all the later tests be perfect. 
, I'm going to have sweet dreams now that I wrote it off my mind (something exciting with bond wires I presume 
)
(cue ♫taps♫ here)