So ... tonight I did a second approach:
I measured the energy, going into my SR/F after first time showing 100% SoC, but still charging.
Until the bike shut down on its own, 1.8 kWh went in.
I'd be interested, how much loss is to be expected just by AC/DC-conversion, and how much by balancing due to 'by burning off excess juice with resistance' (according to MVetter).
These two factors not taking into account it would mean, that, when my SR/F shows 100% SoC the first time, it actually only has
84% SoC - calculated with 11.6 kWh nominal capacity
86% SoC - calculated with 12.6 kWh nominal capacity.
That's pretty close to what I calculated with my first approach.
For the interested:
Up to the first time showing 100% SoC I charged 3 Phase 16 Amps 230V.
After that I changed to 1 Phase 16 Amps 230V (because my 'juice-o-meter' only works with 1 Phase).
What I still have to do, is, to stop charging the first time 100% SoC is on the display, and to look the next day, if the SoC is still the same.