I've just completed my own crude D10 GDP runtime test on low with an Eneloop. I got a total of 55 hours 6 minutes. When I recharged the battery, the charger said it put 1866 mAh back into it, which is a bit below the 2000mAh the battery is supposed to have.
I had charged the battery the previous day and let it rest overnight before starting. For what it's worth, the initial voltage (with no load) was 1.426v, and 0.878v at the end. I also measured during the run and all results except the first and last were between 1.3v and 1.2v. It varied a lot. I suspect it depended on the ambient temperature. My house can get pretty cold. In fact the first time I attempted a run I noticed the light went out during the second night, but it revived the next morning, and I think it was just too cold for the battery. I had to abandon that run because I didn't know how much time had been lost - I don't have any clever monitoring equipment. For the second run, I switched it off overnight, so the run wasn't continuous, but arguably more realistic for all that. (Off times are obviously excluded from the total time.) When it finally went out for real, I tried warming the battery in my hands but that only got me a few extra minutes. When it went, it went quite suddenly.
I have seen people claim that their lights are below 1 lumen on low. I reckon mine is brighter than that, but probably less than 3 lumens. The runtime on high was about 56 minutes, so it lasted about 60 times as long on low. This would be consistent with high being 130 lumens and low being 2.2 lumens, if it scales linearly. (I've not noticed any whine on my unit, but my ears are quite old.)
Chevrofreak
here gets 100 hours from a 2650mAh Duracell and a D10. That's quite a bit longer than my run, even allowing for my lower battery capacity (I would have got 70 hours rather than 55). I don't know if that's because the GDP is generally less efficient, or just natural variation between individual LEDs. Maybe his is dimmer than mine? Or would it be more efficient = brighter, at least on high?
Anyway, I am quite pleased with 55 hours. It's longer than the 40 hours advertised. I'm a bit disturbed by the Eneloop failing overnight - it can't have been
that cold in my bedroom. I really wish now I recorded the temperature, but didn't think of it at the time (I'm not at my best at 3am) and it's not been so cold since. Eneloops are great, but I am using Lithiums for my back-up spare batteries.
Later... I just put the light in the fridge for a while at 4°c, and it worked fine, so maybe my bedroom got colder than I realised.