There is no lumen rating standard among manufacturers. Overseas companies often inflate lumen ratings by 100% or more simply because there is no industry regulation on such marketing claims.
SF's rating system is interesting. From what I can gather, their lumen ratings for their incandescent lights are an average based on the runtime of the batteries. So, if SF rates the P60 at 60 lumens, that is an average lumen rating of OTF lumens for the 60 minutes it is running on the batteries. The P60 may start out closer to 100 lumens and finish closer to 30 lumens, but the average will be the number SF advertises. This also explains how a regulated light like the A2 which has consistently been tested to be closer to 80 lumens is advertised by SF as 50. SF says the A2 runs for 1 hour, but it usually runs about 45-50 minutes in regulation, then drops to moon mode for a few minutes. The average between 80 lumens regulated for 50 minutes and 10 minutes at maybe 15 lumens gives SF the average lumens of 50 that it advertises.
Lumens factory is probably different in that it measures the lumens from the bulb itself, and probably only during the first 2 minutes if using a battery power source. This could give as much as 75% higher lumen ratings than SF's method.
Is one method better than the other? I think SF's method is harder for most customers to grasp, while most other manufactures' lumen ratings are over inflated, or just based on lumens straight from the bulb or LED, so you have to take 30% off the top just for losses from reflectors, glass lenses, etc.
Bigchelis posted the link for some independent testings for these two lamps that illustrates my point.