It is because of the fact, MAP based S.D assumes, a one shot deal. Whatever values are used in the VE table are only "right on" and since they are hardcoded, do not know what the airmass really is and cannot tell how humid it is and we know why nameplates and EPA use a MAF

As the tests above showed simply 1 day difference in weather effected the airmass and if I had changed the VE table with the values reported on day one and then tested on day 2 the values would be off again

Thus for perfect calibration with SD you have a live long task of changing the VE table values and flashing PCM with those changes all the time

GM tried using S.D / VE for a couple of years in the late 1990s and found it was a mistake and went back to MAF for there own reasons but also that EPA via smog rules require a MAF.

VE cannot be trusted as there is too many outside changing factors and cannot assure a car that runs well at one location will function the same 30 miles further down the road

The answer then is unless everything stayed the same and the VE table was made correct then if wanting a perfect tune and AFR it cannot be done with SD/VE

I have seen cases where Corvettes tuned for SD and car was used for track racing that the car's performance ramped in how it performed depending on weather and elevation

Now add the other issues as to the O2 sensorsm, so lets also looke at them and what effects them at:

http://www.teamzr1.com/ubbthreads/ubbthreads.php?ubb=showflat&Number=2550#Post2550