The optical characterization tests are giving some stable data with SR aligned, followed by SR misaligned, which can be used to test if the optical gain of BS compared to DARM changes between the two configurations.

Figure 1. Shows one of this test, accroding to Hrec in both cases (BS and NE/WE) the optical gain increases by the same amount (factor ~1.4).

Figure 2. Looks directly at the height of the BS, NE and WE calibration lines https://wiki.virgo-gw.eu/Commissioning/Calibration/HistoryOfCalibrationLineFrequencies . They all change in amplitude by the same amount, which confirms that there is no issue with Hrec.

This would mean that an optical phase noise in MICH cannot explain the mystery noise, as otherwhise the mystery noise level would not depend on the SR alignment. In particular a displacement noise of BS cannot explain the mystery noise, nor a phase noise inside the CP (change in optical length for S-polarized light). Note that this doesn't provide any information of whether P-polarization can or cannot explain the mystery noise, so it doesn't add information on for example fluctuaing birefrigence causing some of the field to be rotated into P-polarized light.