iirc it was pretty well documented to be the synth playing back at the wrong sample rate or something to that effect.
Edit:
Well I read through your discussion. The easiest way to tell if the synth is "out of tune" is just to find another version of the song from the same tour and compare. Tuning songs up/down by a matter of cents (not a full semi/half step) is pretty common practice to help singers reach notes they can't reach anymore in the original range, so it's entirely possible that they either played a wrong track, didn't remember the song was supposed to be tuned differently, etc...
second edit: did just that and listened to another version from '07. Synth is in the same key ¯\_(ツ)_/¯
The thing is, if there was a sample rate mismatch there would be a tempo increase along with higher pitch. That didn't happen here; the tempo is pretty much the same as the original at 130 bpm. In this performance, the synth is playing
exactly in C-sharp major as my second isolated synth track shows.
I also documented in the thread that I tried purposely causing a 44.1 / 48 kHz sample rate mismatch in Audacity. This raised the pitch far beyond a half step, almost to D major, in addition to the aforementioned tempo increase. So because of the above two points, we can put the sample rate issue theory to bed.
You're correct that keys are often changed slightly to accommodate singers. I do this daily for my students. But to make it higher in this case for DLR when you can hear him straining in his upper register doesn't make any sense. I maintain that someone accidentally transposed the synth up a half step.
Would you please link the other version from '07 that is in the same key?