Relax a bit Harrry, SicSe is pretty much right on this one.
The MAF sensor is simply a sensor that figures out how much air is flowing through it in a way that compensates for both the volume and density of the air coming in, so they refer to it as measuring the mass of the air coming in.
The PCM/EEC, whatever you want to call the processor, lol, recieves an output voltage from the MAFS and compares it to a stored transfer function table to determine how much air is actually coming into the engine. In all closed loop operation the amount of fuel used is mostly determined by the signal from the MAFS. Other smaller factors in fine tuning the fuel are the O2 sensors and coolant and air temp sensors.
The transfer function stored in the PCM is an exponential function, which in this case means that the resolution or accuracy of the MAFS is much better in the lower to middle flow rates, and less accurate as you get closer to the maximum capability of that particular MAFS. Most cars, including the CSVT, are designed so there is a lot of extra headroom in the MAFS so that it never gets into the less accurate region of the sensor, in addition, at higher rpms and flow rates, the PCM switches over to open loop mode and pays much less attention to the MAFS signal, switching over to an open loop fueling table based on rpm and load calculations. Even on my car, which is a pretty solid NA 3.0L with a stock SVT MAF, I don't reach the flow rates that would have me worried about the MAFS being unable to read accurately.
In regards to tuning, a MAF sensor is merely a sensor, an input to help the PCM figure out what it needs to do. Also, the stored transfer function of the MAFS is only strictly accurate when you are using the original stock intake tract, throttle body and everything. Anything that can alter the flow distribution in the intake tract, such as an open element filter, a CAI, changing tube lengths, or even throttle body size or location has an effect on the accuracy of the MAF sensor. Most of these aren't huge changes, but they do have an effect on the airflow velocity profile through the MAFS, so consequently they affect the output of the sensor. These changes can be compensated for by modifying the transfer function stored in the PCM, by a good tuner using a proper wideband O2 sensor either on the dyno or possibly road tuning. Devices like the "MAF Optimizer" alter the output signal of the MAF sensor to compensate for the changes instead of doing it in the PCM. While this works for the most part, there are a number of calculations the PCM makes based on the MAF signal, and it it has been artificially altered, the PCM no longer has an accurate picture of what is actually coming in, and it can affect the driveability of the car. The same reasons for not using a device like a MAF optimizer apply to things like gutting or "optimizing" a stock MAF, though even a few more reasons exist not to do that. Just don't do it, nothing good can come of it. Even aftermarket MAFS aren't that great, quality control is spotty and sometimes you just plain can't get them to work right on a given car. IMNSHO opinion, stick w/ the stock Ford MAFS, if you really need more flow, get an 80mm MAF from a Lightning or Cobra R, or even 90mm from a later lightning or 03-04 Cobra, and have a good tuner change the transfer function in the PCM accordingly. If you have an aftermarket MAF already, and it works good, great, don't touch it, but if you are considering one, save yourself some potential headaches and even some cash and avoid them.
Oh, and one other thing I should mention, even though its not nearly as common among contours as mustangs, but you should NEVER use a "calibrated MAF" to change injectors. Some places deliberately alter the MAF output by an amount to compensate for larger injectors; this is a hack way to compensate for larger injectors, and a load based PCM (which all Contours use) will have some serious issues with it.
