Updates on M31 219BLLL run-time test: the light ran all night on an Enloop AA (standard, not the hi-cap ones), 7 hours later it was still going. I checked the output against an HDS Rotary XP-G2 to mesure lumens.
Initially the 219 was at level 17 on the HDS (so 11 lmns)
7 hrs level 13, 3 lms (this sounds like quite a drop, but to the eyes I couldn't tell a difference until checking with the HDS)
7.5 hrs between level 12 and 13, so the 1-2 lmn range. Took a reading on the AA cell: .923V resting and .62V under load (ie. being used in the light). The current draw at this level reads 50mah. I will continue the test into the sub-lumen range for curiosity's sake but seem to have found the tail-end of a long night of slow declining direct-drive goodness.
8 hrs level 8, .4 lms. Decided to compare to Low-Low on a 1-2V 1xAA MDC using a NiMH (Gene says that should be called .9 lmn), noticeably dimmer. Compared to an M60 running on 2xAA NiMH – nearly identical level.
8.25 hrs level 5, .1 lms.
8.5 hrs level 3-4, ~.05 lms. Stopping the test, cell reads .84V resting. After turning the light off it will not turn back on. Cell will still work in MDC 1xAA 1-2V on low-low.