Lumens is a measurement of all light produced; hot spot, side spill and random light as captured in an integrating sphere with or without optics. Candela or foot candles or candlepower is a measurement of the maximum intensity of the brightest part of the light produced at a distance of one foot normally using some type of optic. Lux is the same measurement of intensity at one meter, (39.4 inches). A good quality fc meter will cost abut $100.00 to $250.00 USD, an integrating sphere and equipment can start at $5,000.00 and go up. Having a test done by an outside source can be $500.00, sometimes more or less depending on who does it and the quantity of lights tested. Anyone can check the output published in fc inexpensively with a low cost meter.
The published results of a manufacturer of optics, (can't remember the exact numbers), indicated that with a constant light source of 45 Lumens the output with a 30 degree optic was about 75 fc, with a 20 degree optic it was about 260 fc and with a 10 degree optic the output was close to 800 fc.
When a company states the output in Lumens there is no way to tell how far away an object will be illuminated. When the output is stated in fc a comparison between lights can then be made. Taking that same 45 Lumen light with a 30 degree optic verses that of the 10 degree optic you can now make a comparison using the fc measurements of witch one will meet your requirements.
LED manufacturers rate LEDs with built in optics, (5 mm as an example), in fc or mill-candela while non optic LEDs are rated in Lumens. The major manufacturers of flashlights began to rate their high performance lights starting with the Luxeon I, in Lumens directly from the LED manufactures data sheet. That has caught on as a defacto standard for all sizes of lights. That method only tells you how much light is produced and not what you can do with it or to tell if it meets your requirements.
Optics and beam pattern can be more important than raw Lumen output. Testing ultra high power lights that exceed 300 Lumens presents another challenge in that fc measurements can be inaccurate at that short of a distance due to the beam pattern. Then you have to go with Lumen output and knowing what the optic dispersal pattern is in degrees.
Curt