In order to justify your stability, you need data to submit to the FDA (they don't take your word for it). When doing your initial stability studies, you try and collect data for the most ideal conditions - no refrigeration needed, stable for X years from room temperature to 30C (to account for conditions during shipment).

No doubt Pfizer lined up a number of early stability studies testing -80C, -50C, -20C, 0C, etc. They recognize that needing a -70C cold chain is a logistics hurdle. If they could avoid it at all, they would have pushed for it and had data to back it up.