In this picture you can see there are dimples in the lens retaining ring.
If you get something like a large pair of circlip pliers or a watch opening tool into those, you will be able to unscrew the lens retaining ring.
This is on high. Not the prettiest regulation, but could be worse. But then I have at least three R2 dropins all generically labelled and they probably all vary.
My one seems to give better runtime though if you look at the current draw (Just before I got a light meter), it would appear to be unregulated.
No idea how that works.
2 hours is still pretty good.......
But it does run for ages. On low it should be an appreciable fraction of forever. Not enough patience to do runtimes on low.
Not bad at all.....would make a good power outage light indeed.
That's all I could look at was current draw. If it was well-regulated, wouldn't it draw fairly constant watts? In that case it would draw lower current as the voltage was high and more current as the battery voltage declined. But if it is basically direct driving the LED, then I know higher voltage applied to a LED just means the LED takes more current.
I might run a battery down to 3.6V and then compare the beams shots with the fully charged shots I have here. Also I'm thinking these batteries may not be up to full capacity yet. Do lithium-ion batteries need to be broken in? That would explain the shorter run time.
No they dont......as the cells start to age they loose capacity....as where nimh gain.
But the light output is pretty steady despite the varying current draw. Doesn't make sense to me either. After all, the voltage will drop over time during discharge too, so the power will drop even faster. Not a clue how that works. If anyone can explain it, I'd like to know. The only thing i can think of is a "burn in" of the LED as it was essentially new when the runtimes were done and some LEDs do have a forward voltage that drops as they "burn in". AFAIK lithium cells are born with all the capacity they are ever going to have and capacity drops from the date of manufacture - at least for the common lithium-cobalt chemistry which has higher capacity than most other lithium chemistries.
Hey, that's half the fun! 
Sorry I have thing for runtime, must be because I never had enough batteries when I was a kid and I appreciate things that run a long time.
Extremely professional review here brted, extremely informative and great shots. Frontpaged.
I decided to give them all some exercise, it was very, very bright in here last night.
I'll do a low runtime now. It'll probably be a day or so till it is done - it ought to run for 16-18 hours on low.
After 2 hours, output is now down 7%. Quite a long time to go.
Watched battery chargers are even slower than watched kettles - which, of course, never boil. And I have 10 more of them to charge.
Output now down to 86%. Quite a way to go.
Output now 82%. Still going strong.
78%. Not dropping appreciably. Got the 18650s all charged again.
Wow now that is pretty good.......
76% and going strong. There sure are going to be a lot of data points on the runtime graph.
76%
Down to 74% now. Wonder if it'll still be lit in 9 hours time when I get up in the morning?
72%. Off to bed so I'll stop boring you with updates till the morning when a runtime graph will appear. Assuming it has decided to turn off by then. I reckon another 3-5 hours should see it.
Long enough Mr. Admin? Or do I need to get some higher capacity cells?