I have read and read over and over about this, on CPF and here. There's much better info over there. Basically, you will never be "within 15%-20%" with a home setup. The only true numbers will cost thousands of dollars. You can use home systems for "guesstimating" only and since there is no standard for these home systems, then there is no way to control the numbers, or the fact that they are all so far off from each other. The only way to find anything out would be to:
Go to a company that has the high priced system and have certain lights tested. Take your sphere with you to test your setup against theirs and make calculation adjustments to your readings, to get close to theirs.
Then send out those exact lights and exact batteries to others and have them do the same thing with their systems, till you had a base of a few people that were getting the same results consistently. Then it all has to be rechecked and verified once again every time you use the system, before trying a new light in it.
Not feasible and that is why I did not try to make a sphere. Even if you could make one that was close, it would cost thousands of dollars to buy a bunch of lights that have been ANSI qualified and it still wouldn't help, since you don't know what they used to power them at the time of testing, or the conditions of the tests. I don't listen to any of the numbers I see around here, because no one here has the correct setup to accurately test. They never will. If you really want good numbers, then you would have to buy a real integrating sphere and the software and accessories to do correct testing. before I ever believed any of the results. What we do around here is total guesswork and it's never really going to be accurate, no matter how much some people "believe" they are doing it right. Saying joe should cut his results by 15%, or john should do this, because he has xxx meter is just more total guess work and does not help solve the problems.
No one should be publishing their lux/lumens readings from any home meters or spheres, because there is no assurance at all, that they are even close to correct, so it's not good to even post them. Same for throw numbers. It's misleading for others. I don't mean that people are doing wrong on purpose, but it's still wrong. Unless it's lab checked and verified numbers, it's not relevant information. Better off just guesstimating from amps, if true amp readings can be gotten. Mine sure aren't.
Just me spouting off, I do a lot of it any more.