In fact, my car stalled out once turning into the parking lot. In my car, it's most apparent during light braking - the RPMs swing a lot (down to 500, up to 1500), and the car nearly stalls out, lights dim, etc. With bad data, you get a really bad idle, because in neutral the engine is basically free to rev up and down very quickly. However, with a shitty MAF giving bad data, the ECU doesn't know what to do. Without a MAF, you get open loop control, which still works okay. The MAF sensor and the O2 sensors basically measure oxygen in and out, and together form a closed loop control. My idle with a warm engine and AC off (almost no load on engine) is around 4.7 g/s. Rule of thumb is that idle is near engine displacement. You also expect an idle base-line measurement. You expect this data to constantly be changing very slighty, also varying with RPM / throttle (more throttle = more air.) (As you can see, the resolution is 0.01 g/s, and it updates about ten times per second - or at least my OBD scanner reads it about ten times per second.)

It sends out a square wave of a certain frequency that frequency varies depending on the reading. Some (like mine) have a reference cold circuit. It basically uses a hot element that's kept hot by current, and air cools it down more current to keep it hot means more air is coming in.

A MAF sensor tells the ECU how much air is coming in, measured in grams per second. The most interesting image is the last one - as it shows all the various ways in which it's dying. Here's the data that shows that it's obviously in the process of dying. So my car has a dying mass airflow sensor. I thought this might interest you folks.

Corvette mass air flow sensor problems