I used the pydarm cli to update the calibration. The inverse sensing function is much better and the calibration is now more comparable to Dan's fudge (LHO 68719). However, there is still much to be desired. I used pydarm to export the inverse sensing function, gains for the actuation stages, and TFs evaluated at calibration line frequencies for calculating the TDCF's. I did not engage the actuation gain filters. Also, the TDCFs can't be trusted right now. That all still needs to be worked out. 

I'm attaching a screenshot of the PCALY2DARM systematic error taken from a broadband measurement. The brown trace is the systematic error from the calibration that was being pushed by pydarm yesterday and in the days prior. The blue trace shows the systematic error with a fudged CAL-CS_DARM_ERR_GAIN of 0.93, and the gray trace is the current systematic error exported via pydarm. Adjusting the MCMC fit range helped get a better inverse optical gain value (see page 1 of the attached measurement report PDF).