So, all tests about limits, motor commands, and receiver-motor connection didn't bring anything.
That would leave receiver software, or motor as cause.
But you already swapped receiver, which rules that out as well.
This would indeed leave the motor itself, as cause, I think.
I'm not familiar...