The lab rat returns with results!
The light in question is the SkyRay L1, bought from manafont. It is a nice light with great build quality. Sadly my sample looks like it has been dropped bezel first onto concrete since there is a large dent in the bezel
It has a forward clicky switch, noticeable PWM (similar to SR-850), the build quality is really high but that is to be expected. The driver is held in place by a large retaining ring where the head joins the battery tube, and is thus easily accessible. The LED tint is beautiful, a perfect 6000K white with no particular tint. Threads are well cut albeit they are bone dry and in need of lubrication.
Ah, pictures!
Current measurements:
3 cell config. initial readings:
High 1.47A
Medium 1.05A
Low 0.50A
After further cleaning, lubrication and deoxit treatment:
High 1.91A
Medium 1.25A
Low 0.64A
Quite odd, as I would have expected the current to drop instead?
Driven off of a 12.0V bench supply it draws 1.87A, or 22.4W. *Assuming* 80% converter efficacy we get ~3.6W per XML. Assuming, again, a Vf of 3.2V at this power we get ~1.1A If. So it should be around 2000 bulb lumen. It is noticeably brighter than the SR-850.
EDIT: Don't make assumptions! The converter board does not get hot so there is no way that it dissipates 4.5W, so it must be over 90-95% efficient (making another assumption, lol).
2 cell config. initial readings:
High 2.36A
Medium 1.61A
Low 0.74A
Ceiling bounce, note that it favors lights with a more concentrated beam like the 850:
SR-L1 167lux (initially)
SR-L1 203lux (after cleaning and lubrication)
SR-850 143lux
At just under five hours combined burn time it sadly no longer produces light. One minute it was going just fine and then it suddenly went out. I immediately turned it off, but it seems that something in the regulator has given in. Though there are no acrid smells or obvious burns on any of the components. Perhaps it is that large FET with its markings partly scratched off? My electronics repair skills are, sadly, probably not up to finding and fixing the problem.