I read this on your webpage; but that's not the issue. Both the Li-Cor and the Apogee have a non-linear response across the light spectrum. The fact that the results of the Apogee are in close agreement with the Li-Cor shows that the sensors are similar in their response to the different wavelengths; but both sensors respond much less to blue light than they do to yellow. The Apogee sensor loses 100% response before 700nm in addition; but we are not as concerned with the red spectrum as the blue.
My question is more fundamental than if one light meter reading agrees with another. What I am trying to get at is that if the sensor is not recording blue light at the same level that it would green, then it's reporting less PAR on a lamp that is skewed toward the blue. This makes it difficult (if not impossible) to compare lamps that are of different spectrum profile because the output of the sensor will differ greatly depending on the color temp of the light produced (even if it is the exact same intensity).
The Apogee product spec sheet covers this issue; but then whitewashes it with the statement "Fortunately, common light sources are mixtures of colors and the spectral errors offset eachother". It goes further to state "The sensor measures green light (500-550nm) accurately, so it can be used to measure the radiation inside and at the bottom of plant canopies." Niether of these statements apply to the situation of lamp comparison; it is not that simple if you are trying to make direct PAR comparisons between different spectrum lamps.
IMHO there is a possibility that the blue spectrum lamps have much higher PAR than you and Sanjay are reporting; it could be that all of the lamps have a mean luminoscity that is somewhere around 500 PAR (white lamps recorded 'high' blue lamps recorded 'low'). Of course, I could be wrong; is Sanjay using a similar method of light measurement?