Wherever or whomever gave you that tidbit lacks understanding and experience behind engine controls. Consider the following:
http://img27.imageshack.us/img27/4035/7mafsdcomparisonclean.jpg
The data on the top trend line is MAP based, while the bottom is MAF based. The trend line (black lines) are representative of actual engine airflow - calculated from an external equation that is big and ugly with many inputs from external sensors on the engine dyno.
Notice the top data is much tighter to the trend line during low air flow? Hmmmmm, maybe this is why companies (e.g. GM) spend the extra $ to use both MAF and SD models to calculate cylinder airmass?
Better yet, maybe this is why MAF transfer functions are being split into sections based on operation range?
Modern day OEM's offer all three basic models and use them at different times, with variable valve timing and camshaft phasing being the defacto - many have moved to an instantaneous equation: or "parametric" VE. Instead of a predefined array, you have a system response curve that is, for lack of other words, calculated on the fly. You would need tables upon tables to do work with cam phasing/valve timing/etc.
Arguing SD vs MAF is so 90s....I hope you like math -- because you'll be arguing some of the ugliest equations when it comes to response curves.