I took snap shots of all the available cameras when 2.7W lock and 4W lock to see any scattering light change.
Only IFI and MCO cameras showed obvious differences depending on the power change. See the attached pictures. Left and right panels are when 4W and 2.7W, respectively. The IFI camera mainly shows the high power beam dump in the view. The left area of the dump is illuminated stronger when 4W. They may be some parts of IFI scattering the IR lights. The MCO camera shows the MCO mirror at the center and an oplev folding mirror behind the MCO mirror. The left column of MCO suspension is illuminated much stronger when 4W.
They must have very high scattered lights.
The snap shots are tiff files and can be numerically analysed furtuer. Any volunteers?
Snapshots are in
/kagra/camera/images/2.7W/
/kagra/camera/images/4W/
Script to take snap shots of all the cameras
/users/kokeyama/scripts/camera/snap_all_camera.py
Thank you for the investigation, Kokeyama-san.
To do the similar things reported in 9301 and/or 9322, please feel free to copy my jupyter-lab script:
/users/akutsu/play/20190622/Beamprofiling_tiff.ipynb
which was originally copied and modified from Izumi-kun's script anywhere. I'm sorry my one is not a useful python sccipt itself but the jupyter-lab script. But if you know jupyter it is eazy to convert one to another.
Or, maybe I myself can play around the tiff files for my joy 
Thank you! I will report if I manage to try the script.
Maybe it would be interesting if we could attach a gigE camera at the AS port's viewport window...