That's going to be right up against the upper limit for that driver's output voltage (they claim 17v max). For them to really work properly the output needs to have enough voltage overhead for it to be limited by the current as set by the sense resistors, instead of being limited because it's run out of voltage. That's the condition that kills them.
Well, if it really is the same as the 7x, you should get 2.8-2.9A using two cells and 4.1-4.3A using three if they are fully charged. 
Comfy. I heeded your advice earlier in this thread about the possible voltage ceiling using the 5* driver. The thing is, I’m using it instead of the 7* driver because if I don’t try it, I won’t have any use for it at all. I don’t plan on getting any 4 x XML lights, and for three I’d use a buck driver.
Also I made a custom contact board for it. It’s soldered on real good, and won’t be coming off.
I proved without a shadow of a doubt the 7X driver is different from the 9X driver, as comfy is saying. Opening post here: https://budgetlightforum.com/t/-/23851 explains it all in summary. Quick summary:
1 - 7 XML/7X beast driver: high amps, runs nice, but only XML's on aluminum
2 - 7 XM-L2/7X beast driver: amps lower and drop in short time
3 - 6 XM-L2/7X beast driver: amps comes back up - runs nice!
4 - 7 XM-L2/9X beast driver: high amps - runs nice, awesome output!
Thank Tom. It’s good to have some hard data on these boost drivers, and thanks for the tip on packing between the boards with Fujik. I have one curing now.
I finished the light today.
It's a Securitying 5*XML 3 x 26650 torch. I swapped out the emitters for XM-L2 U2 1A's on Noctigons. All springs are double copper braided, and the emitters are wired in series with 20AWG Teflon. The driver is the KD 5*XML boost driver
Here are my tail readings using KingKong 26650 cells.
2 x 26650 Per LED
High - 6.48A 2.59A
Medium - 1.28A. 0.51A
Low - 0.11A. 0.04A
3 x 26650.
High - 7.11A 4.26A
Medium - 1.60A 0.96A
Low. - 0.22A. 0.13A
There's no point in even trying to check visual output as it a really sunny afternoon. I'll wait til tonight.
I stripped the original driver to use as a contact board, but when I checked, there were only two little pads to connect the driver. With the driver sitting on the wires and adhered and insulated to the board with fujik, I though the the hight might be a little tight.
I made my own board out of copper clad double sided circuit board. It shapes really easily with a sanding drum. Anywhere i didn't want copper, I sanded it off. Then i cut a hole in the center, so the spring soldered directly to the driver. To connect the front and back sides of the ground ring, I drilled holes, inserted six 20 gauge copper head pins, soldered and snipped off flush. Works great.
On the driver side, I placed the outer ground on top of the copper clad ground ring I left on my board. I flooded all around with flux and let the solder flow. This board is not coming off! It's also very low profile, and I believe, offers less resistance than running through a board and it's traces, and into a wire and back to another board.
I packed the driver boards with Fujik before it was installed. I should have taken a picture, because I made a very neat job of it.
Only some running time will tell the tale of this driver.
One of these http://www.securitying.com/securitylng-6000lm-5-x-cree-xm-l-u2-waterproof-extended-led-long-flashlight/ ?
Wicked. How does it handle the 20A+?
Yes. That’s the light I used. The stock driver was a next mode memory 5 mode driver that put out a total of 3.3A. In low, the PWM was so bad it looked like an old moving picture.
It’s daytime, so I’m not even going to waste my batteries. I’ll run it tonight to get a real indication of the beam.