My philosophical reading of the new/old play clock rules:
If we have an administrative clock stoppage that requires a ready for play whistle (i.e., not an incomplete pass or play ends OOB), the play clock is set to 25 -- unless the administrative stoppage was due to an event caused exclusively by the defense, then it's 40. If the event involves the offense at any point, it remains 25.
If a helmet came off both an A and B player during the down, I would not try to split hairs to determine whose helmet came off first for play clock purposes since the clock didn't stop *because* of the helmet coming off. You collectively consider all live ball action that may cause the clock to stop when the ball becomes dead -- and I would have done this before as well, my thinking didn't change with the new rule modification.
The only borderline scenario that I don't have a definitive answer for is if B causes the initial stoppage, and then A commits an obviously dead ball foul. For example, B's helmet comes off during the down. As the helmet-less B56 is running off the field, A88 punches him. I believe that now that A is involved, the play clock reverts to 25, but I might be wrong. If A called a timeout during this interval, that would supersede the helmet stoppage and the play clock would be at 25, so I think this would too.