Workprint frames 141780 & 781 are the same frame repeated (GOUT 141780).
That’s what I’d expect his PAL GOUT to look like after applying my sync script.
#PAL GOUT
Mpeg2Source()
AssumeFPS(24000,1001)
DuplicateFrame(141781)
DeleteFrame(68664,68665)
Presumably, Harmy could leave in the relevant SE frame where the PAL GOUT needs the duplicate. (I’ve never checked the frames on any of the SEs, but I assume that they are not missing any compared to the NTSC GOUT.)
But I don’t understand the other differences that you pointed out.
EDIT: I’m working on providing Harmy with a frame-accurate reference for the NTSC GOUT.