Back again !
Stripped the whole movement till the last screw and spring. Demagnetized every single item, screws, bridges, balance assembly, pallet fork, you name it, apart from the main-spring which was too awkward to handle through my demagnetizer, which is build on the same principle as the motor in this internet picture;
Inspected every item, re-assemble the movement, re-oiled cap-stones etc. Let it run for 18hrs, wound it fully and it ran one hour before taking measurements. Ignore the timegrapher jump in DD and some other odd off-line dots, that's interference from the computer-fan and at one stage (CD) me moving my hand over the table top;
Dial Down;
Dial Up
Crown Up;
Crown Down;
I would say this is a HUGE difference and as far as I'm aware off, my procedures didn't change apart from demagnetizing, I didn't find anything wrong and so magnetism seemed to have been culprit.
If that's indeed the case, I can not say anything else that I hugely underestimated the influence of magnetism and for sure will take demagnetizing up as a standard part of servicing a watch.
Considering the watch has just had a full service, the amplitudes look good, the beat-errors look fine (I even doubt the 0.5ms in DU, on the graph it looks better, but I'll re-measure) and the daily-rate spread in the different positions seems acceptable to me. There is still a very slight wave pattern, but compared to what is was ..... that's peanuts.
The chronograph parts, dial & hands can go back on and see how it runs .......
Thank you all for you help and I just learned a great lesson !!
If there are any remarks or suggestions left, I'll be delighted to hear.......