The AA Eneloop (the low-discharge formulation you mention from Sanyo) was tested by an obsessive at amazon.com (http://www.amazon.com/review/R2QVW9U1ISXVGV/ref=

The AA Eneloop (the low-discharge formulation you mention from Sanyo) was tested by an obsessive at amazon.com (http://www.amazon.com/review/R2QVW9U1ISXVGV/ref=cm_cr_dp_perm?ie=UTF8&ASIN=B004UG41XW&nodeID=172282&tag=&linkCode= http://www.amazon.com/review/R2QVW9U1ISXVGV/ref=cm_cr_dp_per...) and found to have about 2100 mA-h of capacity. If you take the NiMh voltage (per http://en.wikipedia.org/wiki/AA_battery http://en.wikipedia.org/wiki/AA_battery) to be 1.25, then a simple conversion yields: 2100 mA*h = 2.1 A*h => 1.25*2.1 V*A*h = 2.63 W*h = 9,450 W*s [3600 s per h] I.e., about 9500 joules. This is on par with the Duracell and Energizer results in TFA. I found this encouraging. This conversion does not take into account the drop in voltage that TFA did take into account, but that the above conversion did not (I assumed 1.25V throughout). But even if it ramped down uniformly from 1.25V to 0V, this would only cut power by a factor of two (and that's very much a worst case).