At 25 W; Reducing the CSOFT-Y gain to 20 and increased the CSOFT-P gain to 30 before going up in power.
Powering up with ETMX RH at 0.45 W and ETMY RH at 0.7 W. This looks promising. The IFO was locked with 47 W of input power for over four hours. The lockloss occured at 1303946273, which seems to be due to ASC. No PI ringups were recorded at this lockloss.
I think the high wind might have contributed to this lockloss.