Around 260-300 OTF depending upon conditions and lense/bezel used according to my own measurements. The XT-E should be around 8-10% more efficient in total output.
[/quote]
The stock setup was:
Low-Medium-High-Strobe-SOS
20-120-360 lumens @ 56-330-1047 mA measured at 3.9V
16-106-360 lumens @ 42-280-1049 mA mesured at 3.6V
At 3.3V I measured 319 lumens on High.
So it seems to be a very efficient LED and it runs cooler than XP-G.
[/quote]
So it drops 40 Lumens as the Vin depletes from 3.9-3.3V.
What is causing the drop in output? thermal warming?, lower Vin?... both? I would not expect it to drop that much if it is running cool.
At Vin of 3.3V, you are getting below Vf at 1A for the LED so the current draw is probably down as well. The driver is taking some of the Vin for itself.
You are spot on, brted. The 7135 are out of regulation at Vb = 3.3V and the current drops.
I think I will have to take this dropin to integrating sphere to see, what it gives OTF.
The output for XT-E is given at 85'C. It has to be multiplied with 1.14 before comparing it to XP-G.
So XT-E R5 gives 143x1.14 = 163 lm @350mA at 25'C.
XP-G R5 gives 143 lm @350mA at 25'C.
I use a spreadsheet with formulas derived from Cree data and get:
XP-G XT-E at 1.0A
Led cold: 356 376 lumens
Led varm: 326 346 lumens *)
OTL varm: 261 277 lumens *)
junction temp: 68'C 65'C *)
*) this is for a typical host (L2)
Because these LED's are spec'd based on 85 degrees, would these be perfect for a low driven triple in a small host?
Or a multi emitter headlamp with next to no reflector creating a floody beam and stopping what could be an issue with the colour fringing beam?
I don't think the spec temperature means anything. It is just a setting at which they measure performance. Maybe some people complained that Cree was evaluating their LED's at room temperature when, realistically, the LED is going to get hot very quickly and it would be more realistic to measure the output when the LED is hot. The datasheets allow you to generate numbers at any temperature you want, it's just the default (100%) value is at 25 degrees for the XP-G while for the XT-E it is 100% at 85 degrees.
So runtime should be amazing with this thing, huh?
Exactly equal to any other LED driven at 1 Amp. An XML, XPE, XRE or XPG driven to 1 Amp would have identical run-time**.
**Assuming a linear driver where Vf is irrelevant. A switch-mode driver would vary the run-times a little according to differences in LED Vf.
PPtk
Correct, and XT-E is 114% at Tj=25 degrees.
I wonder why they changed the temperature? Could it be that the carbide base of these chips doesn't transfer heat as well so they expect them to actually run hotter?
On the contrary, XT-E has a thermal resistance to the base = 5 degr./W and the XP-G has 6 degr./W
I think they started binning XB-D and XT-E at 85'C as a new policy. You can easily make a verification test at that temperature. It is more difficult to set up a test at the north pole 
Got my XT-E drop-in from KaiDomain today. I love that it has 3 modes using the NANJG AK-47A driver. The drop-in has a smooth reflector, but the beam has a definite yellow center. The edge of the hotspot is a little green or yellow. Not sure if Cree will be able to work this out.
Here's the LED:
That beamshot looks a lot like the LED I received.
I wonder if manufacturers like Eagletac suffer from the same problems with their D25 series.
I tried some 3000K warm white XT-E in a homebrew 4x Mag mod that I'm not done with yet. I tried a variety of reflector shapes, sizes and textures and even a few optics. The one and only thing I tried that gave pleasant uniform color (instead of banana-berry swirl) were Ledil Boom Spot 20mm microfaceted reflectors. I'd venture a guess that a Fraen 34mm microfaceted reflector might look good as well, but I don't have any spares or room for them in the light I'm building.
Remember when the XMLs first came out we got horrible color change like that also. Hopefully quality control will come up after a while and we will get more even color dispersion
I would love to see a outdoor beamshot to see how the color of the beam translates to a multi colored environment.
but from what this thread has said, i think it would be better to apply the $8 one of these costs towards another xm-l drop-in.
I got one too and while I was a little disappointed by the beam it produces, I don't think it's all that bad. To me, the beam is somewhat similar the what you would get from an Edison Opto Edixeon LED which like the Seoul Semiconductors P4 and P7 also uses a Cree die (EZ1000?). Anyway, if you're a flashaholic and you don't mind having something different in your collection, still not a bad thing to have an XT-E P60 dropin in my opinion. I find the mix of yellow and bluish-white tint of the beam as a hybrid giving you cool and warm white all in one beam.