I always find those "300cuft per minute" figured from sp to be unrealistic. It takes more than 16 seconds (300cuft/min) to empty an al80 with the valve all the way open an no reg at all.
Do you know why that is the case?
Because without a reg on the valve, the pressure drops in the passages inside the valves and the valve can't flow nearly as much gas at lower pressures.
Put a reg on it however, and it will flow a lot more gas because the entire flow path in the valve is now at tank pressure.
Consider for example the really small internal diameter of stainless steel hoses in fill systems - they'll flow a LOT of gas in a very short period of time if you crank the valves wide open.
----
Now... with that said, you are not entirely off base:
1. tank valves vary in the flow rates they can support, and
2. the flow rate will fall as the tank pressure drops.
Scubapro tests their flow rate with a constant 3000 psi gas supply, and you are correct, a Mk 25 won't flow 300 SCFM at 1500 psi, but the question you have to ask yourself is how much does the flow rate have to fall off before it gives you enough additional time?
17 seconds, 27 seconds, 37 seconds - the point is that you've got a finite amount of time to work the problem before you potentially lose all your gas, ad if you're in a tight passage and can't reach the isolator, you're just screwed. That makes side mount a much better choice in some circumstances.
Back mount valve drills are no big deal in warm water, but I learned a long time ago that it's a slightly different story at depth in 35 degree water with dry gloves, a dry suit with heavy undergarments, and cold hands. In those situations I either dove with the isolator closed, or dove with independent doubles, to avoid the possibility of total gas loss, or at least losing so much gas that I could not surface without using a team mate's gas.