I verified the PRG calculation at LHO, motivated by Aidan's PRG comparison between LHO and LLO.
It's correct.  I've updated the medm screen to reflect what I found in the h1lsc and h1ascimc simulink models.  

I couldn't find in the alog where the PRG calculations were, so I'm reproducing them here.

PRG = Power in PRC/Input Power incident on back of PRM.

The input power on the back of PRM is calibrated into watts using IM4 Trans NSUM, fixed in alog 46590.  
If we trust this calibration, then Power on back of PRM/Requested Input Power = 88%

Power in PRC is calculated from POP A LF IN1.
POP A LF OUT is calibrated into microwatts via a "to_uW" filter module with gain = 4.015.  
POP A gets its power from transmission through PR2 and a 90:10 beamspliter (DCC D1000313).
PR2 has transmission of 228 ppm.

Power in PRC [W] = POP A LF IN1 [cts] * 4.015 [uW/cts] * 1e-6 [W/uW] / (T_PR2 * T_BS)
Power in PRC [W] = POP A LF IN1 [cts] * 0.175 [PRC W/cts]


The "calib" filter in the PR_GAIN filter module has gain 0.1745.  Assuming the calibrations of the PDs are correct, the PRG calculation is correct.

PRG = POP A LF IN1 [cts] * 0.175 [PRC W/cts] / IM4 TRANS NSUM [Input W]