Interesting theory, but I would think it strange that it would have enough blow-by pressure to spike the av bay, but not enough to separate the rocket prior to apogee and have a flight failure.
Was this a CTI motor? The 54s are known to blow-by the fwd o-ring, and it seems that more and more reports of it happening with the 38s are becoming common place, too (I've had 2 happen last year myself before learning about gluing the fwd closure).
@jbrracer, is there a shear pin/pins holding the av bay to the booster fin can? How long is the recovery harness from the fin can to the av bay? If there were a blow-by puff that resealed, you got very very lucky.
You say that you've flown this av bay before, but has it been in this same rocket? The whole bay or just the avionics sled?
Was a motor delay used as backup to the apogee separation charge? It's entirely possible that variance in the delay blew the motor ejection charge at the same time that the avionics charge went off, causing a massive pressure spike that blasted the forward section hard enough against the recovery harness that the nose shook off and you had an un-commanded deployment of the main-at-apogee. Regardless, the descent data looks like the main came out at the top instead of 1K feet like you planned.