I agree this is the key statement :

Another strategy may be to obtain narrower claims that better avoid the prior art yet still read on a competitor’s product.

We have been forced to narrow the claims in order to distinguish the Norris Voice recorder from the Sharp Voice recorder. And that strategy has been to identify the removable flash power control, ie high voltage control using MOSFETs gated with the processor.

The problem is that for many years flash has not required high voltage for programming, and in fact have only required the standard Vcc, at 3.3 volts. So not only is there no advantage to using a MOSFET to control the power, it has a disadvantage in that MOSFETs are leaky, and since designers want maximum battery life, we stay away from that.

But it seems we have no choice. If we don't narrow the claim to something like this, we won't overcome the USPTO rejection. It is quite a conundrum.