Since we can't go all the way up in power right now (Team PI is working furiously so that we can soon), I took some time to see if I could move the SOFT offsets after the PRM had been moved to wring a bit more power recycling gain out of the interferometer. 

The short answer is that yes, it seems like (at least at 20W) we can get a few % more recycling gain by re-optimizing the SOFT offsets after we've found the right spot for the PRM. 

The longer answer is that it's tricky.  Maybe this will go away once we go up in power, but right now I'm seeing a pretty high peak at the 28Hz that Sheila pointed out the other day, that is probably the PR3 bounce mode.  It increases as I improve the recycling gain by moving combinations of PRM and Soft offsets.  Eventually, suspensions start saturating and I see large harmonic peaks in the DARM spectrum - clearly this is no good.  Since we weren't seeing this peak once we're stable at 40W, I'm hopeful that it won't be a problem, and we can get the extra few % of PRC gain.  But, we probably won't know until we give it a try.

Note to self: minus for DsoftP, plus for CsoftP was the good direction (this moves Xarm only).

I don't know that it's super helpful, but attached is my StripTool showing the inputs of the Soft and PRC1 loops (so minus the respective offsets), and the resulting power recycling gain.

Also attached is a plot showing that the 28Hz mode is in all of the ASC loops, so however it's getting in, it's getting passed around like a bad cold.