With the difficulties locking, I tried changing some of the CPS diff filters to get the end station ISI to follow the corner ISI. The CPS diff uses a band pass filter to reduce the differential motion between the ISI, but this filter can inject motion at the microseism. Microseism is low, I'm guessing the arms are struggling more with the below .1hz motion. In the attached image the filters we had been using, are the red trace, blue is what I switched to. It kind of seemed to help, but I think we lost lock after I left the room. I'll leave this running for now, going to try a couple other things.
These changes were reverted today, in case they were causing issues with locking. ASC was engaged at the time, I didn't see any effect reverting to the old configuration, so I think they were probably benign.