If the control shaft had been keyed rather than a plain rounded shaft the outcome with binding at the TR end would have been different. Of course, this could have moved the failure point elsewhere if still not detectable to vib sensors or feet.
The design of the locking nut and carrier assembly looks totally appropriate for a rod that will only ever act in a push/pull manner. What the designer appears to have missed during failure analysis is that binding at the TR end would transmit the rotational force of the tail rotor onto the linear control rod inside the powered shaft. The rod was neither designed to freely rotate in this failure mode or mechanically prevented from spinning. It does not appear to be monitored by sensors either, so this control rod may have been spinning for a while.