Starting from the noise budget estimations of losses asymmetry around 180ppm, one can have a look at how the CMRF (coupling of common noises to dark fringe) changes with finesse asymmetry, using Francois' equation.
The first two plots shows the CMRF for some different values of finesse asymmetry. The optimal value is clearly negative. The third plot clarify this: it shows the CMRF at 50 Hz as a function of finesse asymmetry. The green dot shows where we are now, the red one the maximum asymmetry we can have.
It seems therefore that the best operating point should be with even larger finesse asymmetries than what seen so far: we should push north finesse to the minimum (B7/B5 = 0.387) and west finesse to the maximum (B8/B5 = 2.136).
Finally, just remember that so far finesse and losses asymmetries computed online are half with respect of the one reported in the noise budget. This calibration difference will soon be fixed.
