10 lumen brightness continues over 3h mark, I just aborted the test there.
During the 10 lumen period it's possible to switch back to high with a single half click. High is then on for 2-3s and steps down back to 10lm emergency mode. This can be repeated as long as the batteries last.
That photo is a bit misleading.. The spring mod is only on battery carrier negative spring. Doesn't help 26650 usage. [Frankly, doesn't help lot in AA usage either - rumor says Hank is even going to remove it from further versions]
It should be noted that it really is negative spring. Positive end of the carrier doesn't have a spring, which fools me to try to put the carrier in the wrong way around. Carrier spring towards the tailcap is the rule with this light.
As I mentioned earlier, the lens is very clear. Re-checking it now in a direct sunlight shows signs of some kind of coating.
It's not as purple as I have used to see in AR coated lenses, but it shows fingerprints "better" than uncoated lens.
So I decided to test initial OTF lumens with and without the lens and the results were good: Lens cuts only 1.6% (13lm) and performs like an coated lens should.
I think I need to admit my mistake: The lens has (some kind of) AR coating and it performs well.
I have a lense just like that the its in the oem d5, no sign on the specs of coating, never noticed for MONTHS it was coated, went to take a picture one day and spotted it, now I don’t know how I missed it.
I used King Kong 4000mAh in my tests and (as unprotected) it worked fine. Now I tried TF 26650 "5000mAh" and that's too long at 68.5mm. Same applies to XTAR 18700.