One possibility without knowing the 'last permitted' deflection is how little rudder deflection is needed to overstress the VS at high mach. Without much in the way of instruments, what's left? Feel. A yaw is uncomfortable and telegraphs its stress in the seat cushion. The 330 is no different. That the Rudder was limited to 6 or 8 degrees doesn't prevent severe stress from impacting the tail. If there was coupling that followed an oscillation, that would be bad. There is not nearly enough of a sequential trail to establish much. A question I would have is how far into and up to the a/p disconnect limits did the a/c proceed? Was it one limit that nudged the trip out? A combination? The most important data is the amount of activity in each of the three axes were the pilots left with at disconnect? Not only that, but how much authority is too much for limited/degraded flight cues from the panel? Boxes.