Only EX shows a sharp peak at around 15Hz.
Arai-san pointed out that this is probably due to the change of the suspension wire thickness from 50um to 100um for the X suspension (Y uses 50um).
This change caused an increase in the pitch resonant frequency resulting in a poor coupling with the upper damping stage. Therefore, now the mirror pitch resonance has a high Q-value.
The second attachment is the simulation by Arai-san which beautifully explain the origin of the peak.
From a curiosity, I also checked the spectrum of REFL PDA1 with only Y beam. There is no 15.25Hz as expected, but there is 17. 25Hz. There is a coincident peak on BS oplev PIT. Note that BS oplev QPD is not diagonalized yet, and PIT doesn’t necessarily mean pitch with respect to the horizontal axis.
That about BS oplev diagonalizaion is true, and don't forget that the BS's coil balance would not be tuned well either, and will be done by MIF during the break time in the beginning of April.