Fast forward 8 weeks later…
My apologies upfront for the long message - maybe this helps someone in a similar situation.
The batteries are what I would call successfully revived. Every time I recharge I have put the bike outside under a separate roof where a fire cannot cause further damage. The diagnostic tool shows a good balance and the bike does not loose an awful lot of charge over time (circa 2-3% SOC over 3 weeks).
The bike is registered so I can ride it on public streets - which I do. And found a strange misbehavior:
It appears to have *VERY* little power from start. Hard to guess, but I would say probably not even 1kW. On a traffic light stop, even 50cc scooters are much faster. I am slowing down ALL other traffic. But as it gains speed it also gains power and at the end it does reach about a 100kmph - just as advertised. The dashboard also indicates current power and it supports what I observe.
Also, the diag file (the bike has a USB drive where it constantly records all sorts of variables) shows high motor temps up to 110degC and high controller temps up to 85degC. This leads to a thermal throttling message on the dash board and the bike is limited to 65kmph.
When I stop & measure both parts with an IR thermometer they are below 40degC.
So this indicates something‘s wrong between batteries, VCU, controller and motor. Maybe SW, maybe HW/electronics.
The batteries still have errors in the diagnostic tool, but I assume this is not the reason as the errors are about battery block imbalance (30mV allowed, my bike as 45-50mV) and temp sensors sometimes being off (4degC allowed, my bike has 4.2gedC). Also, the errors only show in the diagnostic laptop SW and not on the bike dashboard I again I assume they are not super critical.
I have started with the controller. I got my hands on a spare bike with one of two batterie blocks still present, the controller, a DC-DC and the motor. Not to mention the mechanical parts like fork springs, wheels, brakes… But no VCU.
I swapped the controller - no effect at all. Temps still go up and the VCU itself is still below 40degC - same for motor.
Next, I was luck to get my hands on a spare VCU. I was almost super certain that the error is with the VCU as the bike wiring diagrams do not show any temp sensors so when there is a too high controller temp reported it compels from a sensor on the VCU PCB. Conntected it and besides error codes indicating that the real time clock is not working (backup battery was dead, meanwhile replaced) the bike again works just as before

- apart from changed ODO.
Obviously the VCU was not the root cause. Guys I can‘t express how demotivated I was after that experience…
This leaves me with swapping the motor as last chance. There is a very small chance that the encoder is misaligned which triggers all sorts of strange behavior. I can‘t see the connection to high VCU but I do have the imagination that low power at low RPMs might be in connection to the encoder.
And this is the status of today. I do have a little time over this weekend but it is also a hell lot of screws to dismantle and definitively not done in an hour.