This video shows the movement of the green beam on the target at 290 m when the local control loops of the suspended injection mirrors (PR and BS) are closed. It seems to me that it is moving too much in the vertical direction. 

In the attached pdf there are the spectra of the PR and BS mirrors.

When the loops are closed, there is still a narrow peak at about 7.5 Hz in the pitch of the PR, that is cleary visible also looking at the loop error signal in time domain. According to what observed here, I would say that it is due to the intermediate mass touching the magnet holder plate. In the past we observed that it happened often after moving a lot the yaw picomotor. 

We should now try to understand to what extent the beam motion is affected by this and possibly solve the problem.