The Eneloops have slightly less capacity (@ 2000mAhrs) than many other NiMh cell (2500-2700 mAhrs). Lithium Primary AA is rated @ 3200 mAhrs at a slightly higher voltage so flashlight manufacturers cannot help themselves but quote run times using the highest energy density cell such as Primary Lithium.

There is an excellent article showing the measured run times for the L2D are available here;

http://www.candlepowerforums.com/vb/showthread.php?t=155819

The Lithium AA cells giving an excellent 86 hours. But the measured light output is only 6 Lumens. Comparing like for like 2AA 2600mA gives 60 hrs runtime again at 6 lumen. If we scale for the smaller capacity Eneloops i.e (2000/2600*60) we get 46 hrs. If my Fenix L2D was actually outputting 12 Lumen rather than 6 lumen then it is not unreasonable to assume(6/12*46) = 23hrs usage. (my 22 hour estimate was calculated by measuring the current draw from the flashlight i.e. 2000/90 where 90mA was the measured current considering the excellent flat regulation). Measuring draw current is a lot easier than measuring lumen output I'm afraid.

I have my doubts that the Fenix LD20 according to the Fenix website can actually output 9 Lumen for 71 hours even when using Lithium Energiser AAs. 71 hrs would mean a battery draw of around (3200/71) = 45mA on the lowest output and an output around 6 Lumen, but then again determining whether 6 or 9 lumens (50% brighter) is being generated is very difficult to determine without specialist optical kit. This is probably why subject photographic comparisons of light outputs between known flashlights are so popular on the Candlepower forum.