Summary:
We have 2 dead tiles: WPP6TT1 (EQ3 Bd4 ch13) and WPP6TT1 (EQ2 Bd8 ch15)
One tile which responded with very wide pedestal (+neighboring channels) W PP6TT19 (EQ4 Bd10 ch27)

========
I went again through the 7 patter test from March: link
and checked again all the channels/ tiles.
I made the pdf with all the info about "bad" tiles from the pattern.

I ran the pedestal runs (26157012-29) changing just one vped for given tile and verified with the pedestals if they changed.
The summary of the findings is on the first slide here

All the tiles responded what means the mapping is correct.
Except of one: WPP6TT19 (EQ4 Bd10 ch27) where when I changed the vped the pedestals got very high sigmas - plots attached.
Run 26157012 - default
Run 26157020 - change of WPP6TT19 vped