Since Alicia and Joan-Rene found that there may be some correlation between some combs and the ITM ESDs, we're interested in trying to ground the ITM ESDs.  But, since we use ITMX during the transition of the ETMX ESD electronics, we need to change the transition such that we use ETMY and not ITMX.  

I started with the LOWNOISE_ESD_ETMY state that we had put in some months ago, and made it look more like TRANSITION_FROM_ETMX.  Having done that, I think we can go LOWNOISE_COIL_DRIVERS -> LOWNOISE_ESD_ETMY -> LOWNOISE_ESD_ETMX, rather than the usual LOWNOISE_COIL_DRIVERS -> TRANSITION_FROM_ETMX -> LOWNOISE_ESD_ETMX.  I've made that a viable path in the ISC_LOCK guardian, but with a giant weight of 15 so that it won't ever take this new path on its own.  Note that this means we don't need at all the BACK_TO_ETMX state, which I've removed from the path.

For changes to LOWNOISE_ESD_ETMY,

• I moved turning off the L3 lock gain to before the DARM output matrix starts getting sent to ETMY; this is consistent with and more similar to TRANSITION_FROM_ETMX.
• I set the ETMY_L1_LOCK_L_GAIN to 1.2 (from 1.06) to match TRANSITION_FROM_ETMX.
• I set the ETMY_L2_LOCK_L_GAIN to 15 (from 23) to match TRANSITION_FROM_ETMX.
• I set the ETMY_L3_LOCK_L_GAIN to 1.25 (from 1.0) to match TRANSITION_FROM_ETMX.

I left the ETMY L3 DRIVEALIGN L2L gain at -34.7, since that's what it was last time we tried this (looks like around March 20th 2025, 16:20 UTC).  But, I think we'll definitely need to double-check that number by measuring the DARM loop.

As Sheila suggests, the right thing to do to test this is probably to do the state LOWNOISE_ESD_ETMY by hand, so that we can go partway and measure the DARM loop, to make sure the gains are kind of okay.  We might give this a try sometime next week.