Btw, your L1 of 0.846 for LED white daylight puzzles me a bit, that suggests a spectral error to correct of 15% and in the OP I did not find an error as high as 15% for any of the white led sources.