I performed health check of PR2 (fig1-33).
IMR TF gain reduces by a factor of 10 though we recovered the IM OSEM position by moving BF/SF GAS (klog29918).
IMP gain becomes smaller but seems fine because the gain reduction is not so large.
TMP gain also becomes slightly smaller maybe due to micentering of OpLev beam.

I'm not so sure why the IMR gain becomes smaller but it might be helthier than before the vacuum evacuation because currently IMR local control has -20dB gain, which means IMR gain was somehow increased from the one before the earthquake.
Further investigation on IM OSEMs is necessary but at least no rubbing seems to happe in PR2 now.

Note:

I checked the OSEM value when measuring in-air health checks (blue reference in the graph).
Figure 34 and 35 show the OSEM raw input during in-air (reference) and in-vac (current) health check respectively.
OSEM signals seems much closer to the nominal vales now.