There are limits and then there are really limits.
The real limits are defiuned in the limitations sections as regulated.
Thus the limit on mechanical speed is based on demonstrated min stress margins both centrifugal and vibration.
The temperature limits are based on time and temperature. Up to a certain level time is not important within the expected life cycle. Beyond this level time has an increasing bearing of accumulated hours or minutes.
Overboosting by command (within the other limits controled by the engine itself)is a rare contributor towards exceeding temperature and time to the point where a failure condition is imminent.
The greater risk is where an engine is allowed to deterioate by wear to a point where above average maximum temperatures have occured over thousands of flight cycles (unregulated and often untested)as in a case where maintenance is deferred until failure)
If you now combine such a deterioated engine with a commanded overboost (limited only by RPM) of course a failure may be expected. Fortunately the data is quite scarce in this regard