This will likely be a source of debate but in my opinion you are indeed correct in saying that metering every bit of air would indeed make for an accurate a/f ratio. However with our cars there are limitations to that, namely the MAF circuitry in the ECM. Really there is only 1gm/sec accuracy for up to 255gm/sec, to go to 512gm/sec that accuracy goes to 2gm/sec and to go to 768 that accuracy goes to 3gm/sec. Also the MAF meter itself can have a slow response time in WOT throttle situations.
For part throttle (cruising etc), a MAF type system would likely give you better fuel efficiency and driveability since the amount of fuel delivery is calculated based on an actual measured air delivery.
Keep in mind too that the ECM in our cars was designed to behave with no closed loop WOT, add to that the narrowband O2 for part throttle and it relies more heavily on the MAF reading.
Now enter speed density and consider that it makes it's fuel delivery decisions based on Manifold pressure (MAP) and manifold temperature (MAT) to calculate the Volume of air in the engine (this is oversimplified of course) and then makes fuel adjustments accordingly.
This makes for a robust, fast response system but is somewhat inaccurate, now if you add the a wideband O2 (an accurate a/f measure) to do the short term trim adjustments and you can have the best of both worlds.
This is my understanding, I am sure I may have some points off, but I think I am at least close.
Also in direct response to your question, we had a PT67 with GN1's at 25psi boost and it flowed in the low 600s gm/sec.
