Tks Fabulist for this input. That is very impressive how the engineers of Fisker did it wrong. That remain very basic mechanics, calculate the torque, the HP, acceleration and shear strength of metal parts. Rotation and torque are common stuff in several applications such as pumps and impeller.
Well, I will put all that out and check for the splines as well. If I see a problem, I'll try to find a way to weld or clip all that together instead of leaving some potential free space.
However, according to what I saw on previous post, noise on decel is more associated with the drivetrain. That is what I noticed before it broke. So that make sense to believe it is the drivetrain.
The post I read was from 2013. So far do you know if some found alternate ways to prevent failure, other then replacing the motors (another 8 000 $ I think)