You can parameterize the problem to bracket the solution. It would help, however, if you knew the drive current for the drop-in. But let's assume the fairly typical 1A drive. You should double-check my calculations and methodology below since I'm doing it on the fly.
Vf could be between say 3.5V and 3.7V.
This link says 3.6V.
Thus Pled could range from 3.5W to 3.7W, and we might assume 3.6W.
I typically assume an LED efficiency of 20%, which means waste heat from the LED of about ~2.8W to ~3.0W, with our assumed value equal to ~2.9W.
Driver efficiency could range from about 65% to 90%. The driver in the DX6090 has an efficiency toward the lower end (about 70%) and has a reputation for relatively hot operation (it is a Cree XR-E P4 drop-in, in fact). I've measured close to 90% efficiency for some of the drivers from The Sandwich Shoppe. 80% is sort of a target value IMO that represents good efficiency. So let's assume that the Solarforce driver is 80% efficient (but that our uncertainty ranges from 65% to 90%).
Thus, Pdriver ranges from a low of ~3.9W to high of ~5.7W. Our assume Pled combined with our assumed driver efficiency gives us Pdriver = 4.5W.
The waste heat from the driver thus ranges from ~0.4W to ~2.3W, with our assume value equal to about 0.9W.
Thus, total waste heat could range from about 3.2W to 5.3W, with an assumed value of about 3.8W.
Put your hand on a 4W night light incandescent bulb. That's about the amount of heat you might expect to generate.
I've run my DX6090 for several minutes continuous in the course of bench testing it on a bench power supply. During that time, the drop-in got very hot to the touch (the drop-in was sitting in air, out of the flashlight body). This could be comparable to an all-Nitrolon environment, where there is not much to transfer the heat away from the metal drop-in (actually, the Nitrolon case could be worse, since there is less convection to help cool the drop-in).
I've done some thermocouple measurements on other drop-ins and when the perception is burning hot, the surface temperature is usually over 130F (54C), and probably hotter (especially if you are up at the high end toward about 5W total waste heat). When you then factor in the junction to solder point thermal resistance of 8C/W, the LED's junction temp is about 30C hotter than the heat sink temp. IMO, you'd like to keep the junction temp below 100C (preferably below 80C, where Cree specs the XR-E's 50,000 hr lifetime, but I don't think going to 100C is a big deal for a flashlight application).
I think you'd be close to or possibly over 100C junction temperature in an all-Nitrolon setup. I personally would do like SureFire does -- swap out the Nitrolon bezel for a metal Z44 bezel. Otherwise, use the light only for intermittent bursts.
However, if you are going to use the light rarely, then perhaps an LED lifetime of even 500hrs or 1000 hrs is more than good enough. In that case, then you might be ok to use the light in any fashion you please in your all-Nitrolon setup -- continuous runs (I still wouldn't make them excessively long though), short bursts, whatever. I doubt you'll exceed the junction temperature limit of 150C.
![Pampers Sensitive Baby Wipes, Clinically Proven, Fragrance Free, Unscented, Water Based, Hypoallergenic, Multi-Use, Hand Wipes 336 Wipes Total (6X Flip-Top Packs) [Packaging May Vary]](https://m.media-amazon.com/images/I/41KLBHu9CAL._SL500_.jpg)
