You speak about oxide leakage and electromigration - of what you deem to be "common wisdom", which was gathered during the zillions of man-years of research in the field (and resulted in consequent improvement in technology, equipment and processes). And the knowledge of failure rate, max. operation temperatures and alike is a consequence of these.

None (or infinitesimally little) of this can be revealed by the "cook until it dies" method.

JW