I think this is where I will start. I will then add complexity as needed. I like starting designs with the simplest possible solution first.It might work if you applied a small voltage (value experimentally determined) to the motor to maintain tension while it is back-driven and then full voltage to retrieve the line.