You are right, Sorry to be rude;
By all means, Welcome to the BLF!
Enjoy your stay, try the veal, etc.
Stay calm, don’t be alarmed, it’s just a holiday back at the funny farm.
Peace out.
Keith
No worries. Totally at peace. Enjoyed your poem too.
Ha, I like this guy. Let’s try not to scare this one off.
Awww! No need to worry about it. I'm not skittish, at least not anymore. The grey in the hair and in the beard has hardened me to all sorts of initiations. But thanks for the vote of confidence. 
One way people have improved the accuracy of measurements in the community is to build light integrating devices which give pretty consistent readings regardless of the beam shape or angle of the light being measured. For example, if a laser-like light gives the same numbers regardless of which direction it is aimed in an integrating device, then that device probably has good integration qualities. One person who has done a fair amount of work in this area is djozz, who commented above.
Another way people have improved the community’s accuracy is by distributing reference lights which have known amounts of output. Basically, measure a light in a high-quality reputable integrating sphere, note the numbers it produces, and send that light to anyone in the community who wants to calibrate their own sphere. Then measure more lights in the same original sphere, so that many people can calibrate their spheres in a way which agree with each other. Ideally, also measure a variety of both floody and throwy lights in the original high-quality sphere, so other people can make sure their devices are accurate regardless of beam type. Maukka has done a lot of this work, and greatly improved consistency in the community.
Similar to that last bit, some have also taken trips to officially-calibrated spheres used for standardization purposes, and/or purchased calibration reference lights from standards agencies, to ensure that our measurements are similar to the standards.
All of those things help, but for some types of measurements, none are actually necessary. For example, if someone tests a light before and after modifying it, the relative difference between measurements can be very precise even if the absolute numbers might be skewed. So we can get numbers for things like “this lens allows 99.5% of the light through, but that other lens only allows 97.7“. Or ”this ZL SC52 makes 5.5 more lumens than my other ZL SC52”.
Perhaps more to the point, it’s also possible to detect things like when one brand uses a different definition of a lumen than another brand, because their numbers are consistently higher or consistently lower than expected. There are at least two different known “lumen” values in use by flashlight companies, and they differ by about 30% or 40%.
… and there are also rather a lot of flashlight brands which make up numbers for marketing purposes, with little or no relation to reality. Those are usually easy to spot, even without measuring anything. Like, if a light uses a Cree XR-E emitter and claims to make 800 lumens, it’s simply not true.
For a true 100,000 lumen light though, all of those so far have quite a few LEDs. This actually makes things more precise and consistent, since the individual variation tends to average out to produce a much more predictable result than one could get with a single LED. There are still plenty of other sources of variation, like the driver components, the battery type and state, the amount of dirt on the contacts, and even the temperature… but at least the LEDs can be a bit more consistent.
I hope this answers some questions.
Yes, it does. Much appreciated. My vault of knowledge increases daily.
You can do what some of us do here to get close lumen numbers from flashlights.
Texas_Ace BLF Calibrated Lumen tube / Sphere No math skills needed - Several spheres still available Lumen Tube
WTS: Calibration lights for DIY integrating spheres / lumen tubes - 67 € - Calibrated lights
Adjust the diffuser’s to get the reading as close as posssible with both Calibrated lights.
This will measure OTF lumens.
Well, you may not believe it, but this thread was started because I happened to read various contributors posts which made me doubt the actual output of a particular torch that I own. So, I wanted an accurate, quick and easy do it yourself method of testing the torch's output. Well, it turns out that this is just not going to be possible, so I'm just going to carry on not knowing if the stated output really is as claimed. But thanks anyway - I can't see myself going that deep in to it as I don't have the tools or wherewithall to do most of this stuff.
Oh well, as for me, if l need more illumination and l have one illumination tool ready at my disposal, then I’m fine.
It depends on what you call accurate, if accurate is within 10%, as most of us here accept as a good achievement, quick and easy to do is entirely possible. Provided that you consider some basic DIY easy to do.
Yeah, I got that. 10% or if you can manage it 7%, seems to be the going rate for acceptable accuracy. I was hoping for a much tighter accuracy rate than that. But as the DIY rout won't give me that and it seems even most advanced high end setups may not be able to live up to my expectations, I'll not explore this avenue any further.
^ What you’ll find is that when people review flashlights they often compare them to others that they own. And relatively speaking, the tint, output, and dispersion is assessed. Most of the time people report lights are fairly true to spec, usually a little on the low side. Sometimes they’re way off. But that seems to be happening less these days. The main reason why is when you’re buying from a reputable brand, their reputation is important. If word starts getting out that a flashlight maker is producing lights that fall far behind spec, it’s going to hurt their sales. People will drop negative reviews. And the flashlight community is pretty good about expressing that negativity.
Yeah, I can understand that and when a manufacturer states their outputs were measured according to the accepted standard, I am much more inclined to believe those figures than other figures for the same model presented from other sources of non-standard testing. When you have a brand reputation and image to protect, you have be quite careful and precise about the product information you release. So yes, I have to believe those figures even though I'd really like to check those figures for myself. I haven't made any comparison between my torches not just because I don't possess two torches that could be construed as equivalent to each other output wise. But also because any by eye comparison is next to useless.
However, 10% error on a lumen measurement isn’t the same as a 10% brightness difference because of the Inverse Square Law.
Have a look at this: Inverse Square Law …or Why a 800 lumen light is only twice as bright as a 200 lumen light
On the other hand, I first started measuring output not to compare to the manufacturer but to test my mods. Measure a light for a base reading, regardless of what is claimed, then after the mod see if it is as much higher as expected and if not find out why. So the act of measuring has always been, for me, about testing my efforts for gains. Pure and simple.
The search for perfection, in any format, is a black hole you probably don’t want to fall into. 
Welcome aboard…. (hide your wallet!)