You can predict the PAR somewhat based on the graph, but not that much really. You need the lens type, area, distance and pattern of lamps, this can vary all over the place.
Good points. I shall update the original post to include some of this, but I'll recap it here also.
Distance from the LED to the radiometer sensor was 12" in all of these cases
; the sensor was not moved and the LEDs all used the exact same mounting bracket to hold them steady. I build 3 small wooden holders to hold them steady on their temporary heatsinks (1/2" thick, 1" square acrylic/Corian bricks with doublesided FastCap SpeedTape).
No secondary optics were used in any of the graphs listed above. Only 1 graph that has not yet been posted used optics, and it is a comparison of "with/without" optics on a Rebel 3-up. I am awaiting further data for this one.
Cree XPG and XPE LEDs, Bridgelux LEDs, and Rebel LEDs: Primary optic angle 120 degrees.
Cree XRE- 90 degrees.
Satistronics 3w, 10w, and 20w emitters: 140 degrees.