It basically means if you want 100% accurate scientific information, a light meter's good as long as you get the entire scene in there, otherwise, a camera gets you close enough to get a relative feel of the differences between two or more lights if you don't want to measure it absolutely.