Since I have both meters, I check with both and the drop off between PAR and Foot candles is very close at least for MH.
But all spectrum don't fall off evenly. For example, if you hake a lux reading when the lamp is brand new, then take period reading very few months with the idea that when the lux reading falls by 10% you will change the bulb. Even with a conversion chart, if the green/yellow spectrum changes slightly it will measure much more than if the blue spectrum changes dramatically.

in a multi chip array, each focused led is sending tiny beams(disco) down not a broad wash (mh t5). The Par meter dome receptor a small disk rather then a large flat disk used for indecent intensity readings . so its only picking up the individual led not the overall "incedent" reading. a wider disk would help getting more accurate readings. The Apogee 500 I believe has a wider disk now.

