Accordingly to the commented entry, yesterday I added a notch in the IP-L damping filter at 406 mHz. The notch has a quality factor of 10 and a depth of 40 dB.

• I repeated the measurement and the decay time reduced 40 seconds! Now the decay time with the control system on is 5 seconds shorter than with the control system off.
• The new mystery is the mode in IP-L at 671 mHz. The UGF is 85 mHz and, therefore, the control system should not be damping this mode. However, it seems it does: the decay time difference between having the control system on and off is 81 seconds.

I shoud add a notch at 671 mHz to check the behaviour.