I would like to know without a intergrating sphere, how far does one place the meter from the flashlight to get a good baseline. (I know its not going to be accurate, Just want a rough guess number.) I want to see how much oomphf my home made light puts out using a cree cxb 3590 vs say my plunger light with a bunch of leds that claims 11000 lumen (chinese lumen)
Its not about distance if you are measuring output since any hotspot on the sensor will give an off reading. My advice is if you are in need of a quick fix head to a homedepot and get some PVC pipes the elbow kind and connectors and a end cap and build a makeshift lumen tube, use a light which you know has a fairly accurate measurement and use that as your multiplier for other lights to give you a general ball park for measurements.
If I understand what you’re looking for I don’t think the distance is that important as long as it’s the same for both lights, however you won’t get an meaningful comparison unless the two lights have very similar beams. It’s pretty hard to get an idea of lumen output using only a lux meter.
I would recommend going for an integrating sphere (styrofoam ball) instead of the pipe/tube...thing
For a long while I did the Ceiling Bounce method —- I used the small dressing area in our hollywood bath area — Set the light facing upward towards the ceiling with the meter sensor some where near the light — I found that if I had the light around 1 ft from the side wall it worked best — you need to have a light with known output so you can figure a correction multiplier (your calibration so to speak) I was surprise after I built a PVC lumen tube how close my two methods were
The method suggested in the OP will not give any meaningfull result. But I used the ceilingbounce method for years, it gives fine ballpark numbers, only extreme flooders are measured significantly low:
*Find a piece of white ceiling
*Place the light head up on a surface (say, a table) shining upwards to the ceiling. I always placed it a bit higher, about a meter from the ceiling.
*Place the luxmeter sensor, facing upwards, 10 cm next to the flashlight, at a measured and fixed distance from the ceiling (say, on that same table), for any future measurement of flashlights, use that same luxmeter-ceiling distance or else the numbers are not comparable.
*Measure the lux value
*Find a flashlight that you are confident to know the output (from the specifications or from reviews, or maybe you own a maukka-calibrated flashlight) and measure it using the same method, this is your calibration to convert the luxvalue to lumens.
well this is easy as my ceiling is white, its a textured popcorn though, and a convenient kitchen table underneath. 
I can try this till I have a integrating tube made my only issue is I dont have a calibrating light, my other flashlights are not rated except for my trustfire plunger light, which claims 11000 lumens, guess I need to buy a standard light of some sort.
also, thank you everyone who has replied with information, this is what I wanted. something I can use to get a baseline reading. keep the replies comming!
what you really need to do is build a white pipe or something that reflects all the light to the sensor
calculate area inside the pipe including the sensor
read lux at the sensor
multiply that by area of pipe plus sensor
this times something in square inches, is lumens
the ‘something’ i would have to look up - seems like it is 12.68 for some reason
My stand is that the ceiling bounce method provides better integrated light output measurements for flashlights than pipe-contraptions.
Pipe-builds are conveniently small and easy to build, and that is a considerable merit, but they are not very good in measuring total light output. But good enough is how the world gets around 
the ceiling method will have errors that are a function of the beam pattern
pipe or sphere collects all the light
why would they not measure the total light?
and why would ceiling bounce be better than pipe/sphere?
wle
I actually have 2 kinda known lights, manuf info anyway, the one is a ryobi p717 one plus light, the only measurement I have from factory is 2500+ lumens
the other is a hyper tools 9903 2 in one work light, 1000 lumens
Several reasons, that would be quite a story, but in short: in the pipe, light that goes forward has a significantly higher chance of reaching the detector than light going in an angle (in other words it is a fairly lousy integrator), which makes the pipe very sensitive for beam type. Tecas-Ace’s use of diffusers fixes some of that (on the other hand diffusers introduce extra spectral errors) but it is still not really good, nothing like the integration in a well-built integrating sphere, there is a reason why light people use those. Ceiling bounce suffers less from the errors of a pipe, especially if you keep te light source close to the ceiling and the light sensor a bit further away (if the flashlight itself does not block too much of the incoming light).
How do you compensate/calibrate for different ceiling materials and their inherently different reflective properties?
I found that when using the Ceiling Bounce method —- I had 2 correction factors —1 for throwers and 1 for the High Lumen flooders
With my PVC tube contraption as it is being mentioned and it is indeed a contraption (nothing technical for sure) I put DC fix on the glass shelf I use —that seemed to help make it a lot more consistent—- It satisfices my needs
Firehopper —the 2 lights you mentioned might be ok but I don’t recognize them as being accurate lights
As long as you’re using the measurements for yourself who cares how accurate they are —— I have 100s of lights and through trial and error I’ve come up with methods that work for me
well the one is a light from ryobi, it has 7 cree leds in a somewhat pistol grip shape. if you google ryobi p717 you will see it. and not very accurate, as the output is listed as 2500+ lumens. so greater than 2500 but no clue by how much. the other is a work/area light from walmart. hypertough not hypertools
The materials won’t change usually. If they do, because wife want’s a fresh ceiling painting, the calibration lights need to be measured again. If wife wants a violet ceiling, file for divorce. Or use a different room.
When you say light meter, we assume you mean Lux meter, not a photographic light meter.
Though a light meter can be used with a ceiling bounce comparing output to known lights.
Here is a link to what I did with PVC and a cheap lux meter.
I never got around to doing a write up on the finished version. Need to get off my butt and do that one of these days.
The ceiling bounce method is quite useful as long as you keep the conditions the same for each measurement.
Or you can re-calibrate with a known light for each test.
Or, perhaps the easiest solution, give Texas Ace a jingle and see if he has any calibrated lumen tubes left for sale.
All the Best,
Jeff
Point being I can’t verify your findings becuz I can’t duplicate your unique ceiling characteristics. I then have to take your word for it. Which is ok. I understand your standard reference calibration verification approach. If it works, it works for you.
Butt at least with a pipe measuring system there are material construction standards already existing, ie., White PVC Schedule 40/ASTM standards so I can reasonably/confidently duplicate someone’s readings by a similarly constructed pipe setup as their’s. Then there’s trending consensus using such significantly similar materials. Butt I digress.
PS. If Jeff above gave me his materials list and dimensions and did my own testing then I compared my results to his and came reasonably close to it, that’s a trending consensus.
You can’t faithfully reproduce even Jeffs findings for many reasons. One being the meter used. Another the light to be measured. A third the battery he used, the amount of dust in the tube, the color variation of the material, Jeffs mood, the mood of his wife, phase of the moon and so on.
Why would you want to in the first place? If you need to measure, you need to calibrate. Admit, you’re just inventing the True Luminosity Measurement.