Just looking at the cover to the video, it appears there is a lot more surface area around the head of the light to not reflect (in case of AR coated lens) light back down the light tube. There seems to be a whole lot of variables that causes much confusion. I tried to compare some uncoated to coated lenses using the same host in this thread. Overall, it AR coating seemed to make a difference, but I'm sure there is much I didn't consider in it.
I once built a Courui D01 for a thrower, was quite pleased with the 338Kcd I got from it. Then I got in the UCLp from flashlightlens.com. No other changes, just the lens swap, netted me 365Kcd. Needless to say, I was very happy with that $7 purchase. 
Yea, lenses from flashlightlens.com are real nice and the output increases have been confirmed. Definitely a worthwhile upgrade from the stock plastic mag lens…
I think my modified Olight Intimidator went from 610Kcd to 622Kcd. Then a change to the battery supply got me up to 645Kcd. That big 95mm UCLp is incredible! Invisible, gives the Illumination Machines Lum 5-90 reflector a clear view. 
Agree with other views, it largely depend on the quality of the coating if everything else being same.
I’ve only measured regular (cheap cheap) reflective glass vs no glass, huge difference, >10% in kcd drop in my T13, so your initial lens quality also counts when comparing vs AR.
I’m in the process of building another D01 for a friend, & have bought a few cheap camera filter lenses, which I’m planning on doing a back-to-back comparison for throw readings.
I should have it done tomorrow.
Over here also achieves improvements between 8% to 10% in kcds in my mods with AR multilayer glass lens regarding original lenses without any treatment
if you don’t mind me asking, what mods did you do to your Intimidator to get those levels of kcd? and what model was it? ( SR90, 92, 96, etc.
Thx for the informative follow ups. I have a feeling that the differences netted measured as lumens in my PVC LMD will be lower than measuring throw. Will do a beam throw comparo in the future and remember to include measurements without glass at all.
Again this was just a sample size of one so would be interesting to test additional samples in the future. Looks like a visit to flashlightlens.com is in my future. 
I would also like to know 
Well, the most annoying thing about AR coated glass lens is that the light gets kinda green tint after emitter dedoming, especialy if the emitter is 5700k and under
I had my Javelot default emitter replaced for that reason
Even without dedoming the LED.
That’s why I don’t like AR Coated lenses. I buy 4C or 3B XM-Ls because of the good tint - and then a lens destroys it 
Not for me… I preferably have a bit less lumens but a nice light color than the other way around.
Why not buy a LED that tends more toward red (something with “A” or “D” tint) for use in lights that use AR coated lenses? If you really don’t like the green tint, why are you buying green tinted emitters?
The A and D are more towards pink than red, but I would ask the same anyway. Most B and C have green tendencies at their core due to their Y axis placement. The easier way to expose that green it is to compare side by side with something equivalent from A or D, whoever the famous "dual tint" of the CREE XM-L LEDs can be more or less visible based on the shape and size of the reflector.
De-domed XM-L2 U2 1A and modified driver. The driver makes too many amps for the XM-L, so I removed their “sense resistor” and stacked resistors to get 6.6A to the emitter. *Olight used a bar, like in a DMM, for resistance.
I put a UCL or UCLp in every light that I favor. The UCL glass lens in a MiniMag Light really disappears, well, ok, it does that in most all of em. Without reflections from external light sources the emitter and reflector look really awesome when viewing the light.
I haven’t bought other AR lenses, can’t say I know what y’all are referring to about the green cast…haven’t noticed that in any of mine but I HAVE replaced Olight and other brand name AR lenses that had a yellow tint to the glass. Lay it on white typing paper. Can you see it? You have to be careful where you put down an UCL lens as it’ll get lost very easily!
The green cast thing is an interesting observation. not something i’ve noticed myself with the Ucl lenses.
But it makes sense in theory since the coating’s purple/bluish cast would indicate it is still reflecting those colours to some extent while letting the green/yellow/red through almost completely unhindered?
Correct, when you look at coatings and see a particular color, that's the wavelength that the coating is "canceling" out. AR coatings have no color, here's a great article on the topic:
http://rick_oleson.tripod.com/index-166.html
I should also add that in my testing, I was trying to see the differences in tint with and without the coating (purple) and couldn't find anything discernible. This wasn't merely just observing it with my eyes but I videotaped it and then compared still frames. However, I believe a coating aimed at green wavelengths would have made a difference.
This would explain my dislike of the tint in the nw blf X6. I've been wondering why the color seems really pale and greenish.
As an experiment, try shining the light around with the lens removed…see if it still gives that greenish cast you’re not liking. Curious to see if it makes any difference.
I just tried it. The AR lens of the BLF X6 does indeed cause the beam to look greener than without. Now I wonder why there haven't been flashlight AR lenses that look green instead of the typical blue/purple/pink (at least I haven't seen any). It might make for a more pleasant overall tint.
Edit: I'm curious if the stock non-AR lens changes the color at all. Unfortunately, I don't have one to test.