I forgot to consider the fact that BS longitudinal motion changes PRY length sqrt(2) more compared with PRM since the incident angle to BS is 45 degrees.
Ogaki-kun said that his calibration factor for BS 6.02e-11/f^2 m/cnt is in the unit of counts to meters in differential length between BS to ITMX and ITMY (not meters in BS longitudinal motion).
This is equal to the unit of counts to meters in PRY length from PRM to ITMY.
Therefore, my calculation for PRM acuation efficiency 1.0e-9/f^2 m/cnt should be correct.

However, the expected value for the ratio between PRM and BS will be

PRM/BS = (0.129 N/A / 10.8 kg) / (sqrt(2) * 0.0138 N/A / 18.9 kg) = 11.6 .

Discrepancy between the expected value and the measured value would be from measurement uncertainty or OSEM actuation efficiency variance.