These data from June 2019 can be used to estimate the contrast defect. The previous estimate from Nov 2018 gave 140uW, out of which 10uW is 56MHz sideband.

Figure 1, shows the time series of SDB2_B1_DC power and the optical gain as the dark fringe offset is reduced.

Figure 2, shows the optical gain normalized by the B1 power as a function of the B1 power. This should be a constant if there is no contrast defect. The  blue curve shows that there is some contrast defect, and the red curve shows that it is relatively well fitted with 20uW +/-10uW.

Figure 3, looking at the data after the unlock, there was -40uW of  electronic offset on the SDB2_B1_DC channel at the time of this measurement. So the contrast defect light correspond to actually 60uW on SDB2_B1_DC, and SDB2_B1_DC is equal to half the total power on the two B1 photodiodes. So the measurement is that in total there is 120uW of contrast defect (or other junk light) coming through the two OMCs and reaching the B1 photodiodes. This is similar to the measurement from last year. I don't know how much of that power is due to the 56MHz sideband