With 2 x Solarforce V2 2400s, NCR18650 2900, XTAR 2600, TF Flames 2400, all are 2.01A, batts fully charged (so don't say that you have a better copy because yours take 2.3A)
3-cells, it is 1.34A. So it scales exactly the same.
My meter under-reads by about 10% +/-, so accurate comparisons to your figures is not really possible.
Lux at 8.17m, 213 lux. Trustfire TF-3T6
Lux at 8.17m, 260 lux. Sky Ray SR3800 at 2.5A 2-cell fully charged.
Reflected ceiling bounce lux
UF-980L 79 lux initially @ 3.3-3.4A, drops like a bomb to 60 plus. Ultrafire U80 @ 2.5A steady at 65 lux.
lux 75 lux steady, Fandyfire STL-V6 (expect the same for Sky Ray STL-V2)
lux 130. Trustfire TF-3T6
lux 165. Sky Ray SR3800
lux 255. Fenix TK70.
lux 440. Ebay/Aliexpress "65W" HID <- power consumption is not measured
lux 810 reflected lux. Sanmak SM5200 55W 4300K bulb overdriven with 100W ballast (98 watts ballast input, probably 80W to the bulb). Should be able to do more with the 100W bulb. <- 1.35 million lux @ 1m "throw", corrected should be about 1.6 - 1.7 million lux @ 1m.
Look at the Ultrafire U80. 2.5A 1 cell 65 lux. 2 lights would be 2-cell @ 2.5A 130 lux. Compare with the triple XM-Ls. You do the same amount of light at 2.0A with just 1 more emitter. That's the beauty of multi-emitters.
This scaling is the same for 3 emitters of the SR3800, pump it to 2.5A for 3 x XM-L and it goes linear to 165 lux. <- This is a ~ 20W triple XM-L light.
Compare to the TK70 (30W), it scales the same to 255 lux and mine maintained for a good couple of minutes (usual source for you super long timed lux tests, thaicpf). Good heatsinking, doing 3A per XM-L and maintains the same luminous efficacy as 2A on a lousier heatsink is not easy. (high tendency for the curve to flatten). You might have read form some forummers that the TK70 has lousy heatsinking, no fins etc....seriously I never see anything if they own the light.
Someone is doing a 30 x XM-L on CPF, 9" diameter. Should be fun.
I got 2.3A with my Ultrafire cells, but they are at 3.85V and probably sagging. Not much DMM voltage drop to affect much, we are talking about 12.6V for 3-cells.
No need to clean, everything is way cleaner than the SR3800. Actually it is the same host, SR and TF are interchangeable. Same 4.5mm thick tube, ie solid.
On thaicpf there is no lux (throw) readings, but the pictures show a wee bit less intensity, which is in line with my measurements.
The guys are smart, other than the threads for heat transfer to host, there is also alu ring at the front which transfers the reflector part heat to the front bezel ring/body.
The heat MASS of the drop-in is amazing (it is damn heavy), it gets barely warm, as in less than 40 deg C for a 3 minute run. Ambient temp is 31 deg C here. 145 lux @ 3 seconds, 141.8 lux @ 3 minutes. 2.2% drop. Batteries 2 x 18500 IMRs 1100mAh. I don't want to do more for host -> air heat dissipation, just skews up the picture.
Edit : Cleaned all contacts, same 2.01A current. Think of it positively, on paper it is not going to be as affected by the heat. You can do mid on SR3800 at 0.9A but with PWM. Get the DRY if you need 3 or 4A, CW/NW/WW, but trust me i am not gonna risk gunning even the 3A past 1 minutes here in the tropics and only 2 mins for demo. Remember, that is 3A or 4A 3-cells.
This light does not fit the DBS V3/U80 tubes, so can't do 2 x 18350 IMRs like the Fandyfire STL-V6 or SR STL-V2 does.
Not much DMM voltage drop to affect much, we are talking about 12.6V for 3-cells.