Overall, one of the reasons people prefer to use “3V” (1S) XHP50/70 emitters is the lack of high power boost/switching drivers. Now, taking a look at Texas Ace's XHP50 output test (peak at 9.25A in 2S/6V config) or XHP50.2 output test (peak at 11A in 2S/6V config) it can be clearly understood these emitters can manage much higher current and power.
So, why not some new high power switching drivers? A word of warning, though. We have embraced a new generation of high power linear drivers, I find this great. However, I think “you” went overboard setting the driving current of the new Osram emitters. In the first place one cannot expect any and all emitters to behave like the ones tested, there may be variations. Then, it also isn't wise to deliver full “expected” current/power to the emitter, this being inefficient besides unnecessarily stressful…
For XHP50 emitters in 6V/2S configuration I think a driver with 5.5A or 6A of output would sell, namely to enthusiasts (5.5A/6A is 82.11%/86.05% of peak output for the XHP50 in Texas Ace's test). These would require 15A rated cells to properly operate.
