The M3Ts "225 lumen" assembly is actually pretty close to 400+ *bulb* lumens on a fresh set of cells (depends heavily on brand of cells here). So that's over 250 solid torch lumen. If the Fenix is being rated 225 at the emitter, than expect something like 175 torch lumens best case scenario, assuming the LED is at a reasonable temperature. Reflector losses for LEDs are not as high as for incans because LEDs emit light on a "hemisphere" not a "whole sphere." The more controlled radiation pattern means that there isn't any light being lost "out the back, through the hole, down the tube" like you get with incans. 20-25% conversion losses is more accurate for LEDs, 90+% conversion can be achieved with certain setups on LEDs.
Plus, I am not entirely convinced that "surefire" lumens are equivalent to real lumens.
I am 100% convinced that Surefire Lumens are not real Lumens. Flashlight companies really can not state a single lumen value for a flashlight and call it an accurate rating. The exception being the case of very steadily regulated lights that spend almost their entire run operating within 10% of their initial output. (some lights do some don't, most SF lights have a steadily dimming output, LED and incan)
A Surefire lumen is a marketing placement number that happens to be torch lumen that will occur at some point during the discharge,. It does not occur on fresh cells or dead cells, but somewhere in the middle of that discharge. What I have observed is that some of their products are more heavily de-rated than others and it is purely related to the marketing position they are trying to create with the lumen value situation. (There is nothing wrong with this, product placement is important for consumers, moving the lumen values around to nice round numbers in a certain "order" is totally fine when ALL stated lumen values are in fact lower than you can expect to see)
Obviously, when a manufacture uses a nice "round" sounding number like 225, (be it fenix or surefire) it isn't because 225 is the number that popped up on their measuring equipment. It's a marketing choice to use round numbers. Surefire Took a light that starts off on fresh cells somewhere in the 250-275 torch lumen range, and ends somewhere around 175 torch lumens before falling on it's face, and called it a 225 lumen light. Fenix took a light who's emitter is rated to deliver some range of lumens (probably like 217-230@I) and called it 225 also.
To get an idea how I like to rate lumen values... check out the compatibility chart link in my sig line, these are torch lumen values fresh off the charger, followed by torch lumen values when the cells are about to "die." Personally I wish more manufactures would just post runtime plots, and pay to have an integrated sphere take a measurement of the light with fresh cells (then they could just correlate that to a lux measurement on a simple lux meter to make their runtime/output plots).. Some are starting to do this. We are seeing manufactures and distributors of lights in the marketplace with good charted information about their lights.
M3T max is 225 lumens, I always thought it is 250. 
