I agree that this battery source really shouldn't be used in modern flashlights, due to their propensity to leak and damage the light. But I'm realistic to know that most non-flashaholics still use them, :sigh: so I like to provide some comparative data.
It's true I don't typically do anything other than Max mode alkaline runs on 2xAA, mainly because it just too time consuming. But I have done a lot of low mode run 1xAA tests over the years, and can provide some general comments as to how 1xalkaline compares to 1xeneloop.
In *very* simplified terms, maximally driven 1xAA lights tend to produce >2X as much light overall on eneloop vs alkaline - and of course regulation is infinitely better on NiMH. As you drop down to lower outputs (e.g. Hi, Med), those differences diminish (i.e. regulation improves on alkaline, and runtimes begin to approach eneloop performance).
Although highly variable, by the time you get down to the typical Lo output settings in good current-controlled lights (e.g. ~10-20 lumens), 1xalkaline and 1xeneloop typically produce roughly equivalent runtime patterns for the same output. Below that, it is possible that 1xalkaline could even outperform.
I have limited data in 2xAA lights, but I would expect the Quark QAA-2 to be roughly equivalent between alkaline and eneloop on its Med setting (rated at 18 lumens) and even lower Lo/Moonlight settings, with potentially even an advantage to alkaline at the Lo/Moonlight modes.
But again, I always advise sticking with rechargeables or L91 lithiums in these lights - you wouldn't want to ruin an expensive flashlight with a cheap alkaline cell (and even Duracell alkalines can leak).