I tried the FF filters Gabriele put in  78307: the new SRCL filter was on at 17:36, (while Jennie W was adjusting the A2L) the new MICH feedforward was on at 17:45 with quiet time from 17:45- 18:10.  In this configuration the SRCL coherence was high, so I switched it back to the old SRCL FF (with a gain of 1, rather than 1.14) at 18:12, and this is how we've been observing.

The PRCL coherence is the highest of the LSC coherences, with MICH and SRCL lower than PRCL but still large from 15-25Hz. 

The attached screenshot shows that we still have large ASC coherences with DARM, as well as jitter coherences. 

Edit: Adding a second screenshot that shows the coherence of SRCL, MICH and PRCL with DARM at 17:45 UTC, when the new SRCL FF was on with a gain of 1 (and also the new MICHFF). 

One potential explanation for what happened is that something happened last Tuesday that means we need an offset in AS72 now, but we were running with SRM misaligned last week before we realized this, which changed the feedforward.  Now that we have a hadn tuned offset in AS72, the SRCL and MICH couplings are more similar to what they were before last week, so we better off with the old feedforward filters in.