Quark Pro QP2A vs. QP2A-X maximum run times?

I was looking at the Quark Pro QP2A and the Quark Pro QP2A-X from FourSevens, and one thing that caught my eye was the huge difference in maximum run time. Both of these lights use two AA batteries and have the same minimum output of 0.2 lumens, yet the maximum run time for the QP2A-X is only 360 hours, compared to 720 for the QP2A (even though it's still a lot). What accounts for the discrepancy?

Probably the different led makes a difference, but I'm sure somebody that understands the electronics and such will be along to give a better answer. I'm sure you can go to cpf marketplace and ask the question in the 4sevens sub forum

There's a typo on the summary table's specs - if you click on the detailed specs, you'll see the X is 0.3 lm. I have an older XPG-S2 and XML v1 and have measured them to be 0.17 and .33 lm respectively. I've also run them side-by-side with a AAAA (from a 9V) and got 100 and 50 hrs respectively, so half the lumens ran twice as long. Just for one more data point, on the same scale, I measure my SC52 "0.34" lumen mode to be ~ 0.08 lm and it's spec is 3x as long a QAAX... so they all seem to have about the same lumen-hr efficiency.

The "bright" floody QAAX moonlight is my person favorite for reading and close task work with night adapted eyes - it's become my new "low" mode and probably the one I use most often.

Like -onahill mentioned the emitters are different, the XP-G2 runs at a slightly lower forward voltage than the XM-L(2). That's not the whole reason but I imagine it contributes some.