Thank you all for your comments.
The Motor , i started with a belt around the flywheel and pulling with my arms, just like an outboard motor.. Ok, it´s very hard to pull but finally works 3rd pull.
Sound with or without cover is the same. 
Conrod bearings and main bearings had run 1000Km after overhauling. I have a C4 bearing left side and a C3 right side, so should be ok. Everything turns without any rough feeling..
But: 200Km ago, i had a defect at the oil pump and therefore piston seizure ending in new bore and piston.. The small springs and blocks from the oil pump, i never found anywhere, but dont think, they could have passed the back-pressure valve at the oil pump and the small hole in the carb , to get into the barrel. Hopefully !
Mark T , is it possible to press in the cranck seal in too far? Doesn´t the castle design on the seals avoid to cut the bearings from oil starvation? Or maybe the touch the inner ring of the bearing?
Very interesting idea to let the motor run without charging circuit connected, but how would you do that at the Vape ?
 

If you installed the seals with the castle turret bumps towards the bearing, it should be okay.  Anything is possible but very doubtful that's the problem since you installed the seals the correct way.

What I would do is simply disconnect the two (black?) wires from the stator to the voltage regulator... see if sound changes. 

Unfortunately, with the new info about the seizure, I'm afraid either the low oil situation or some debris from the seizure has damaged the bearing.  It may even be spinning in the case. 

 

After lots of investigations, i found out the source of the noise.
The noise was madeby replacement reed valves, made from carbon sheets.
On a certain RPM, they seemed to flatter.
Switched back to the OEM metal version and all ok.
Didn´t bring any benefit anyway