Actually the screen is there for a reason - hence the stalls experienced with removing it. The screen is actually there to help remove the turbulence coming in through the upstream inlet. It does a rough job of "straightening" the airflow so that the air meter element will read an accurate amount of airflow. With swirls and backflows over the element, an incorrect air mass reading will be measured.
Sometimes it will be too high (which will result in overfueling and a closing of the idle speed control valve) other times too low (and thus too little fueling). These measurement errors will be most pronounced (larger percentage) at low total air mass readings - like decels and idles where you will be ingesting say <1 pound per minute of air. This is attributed to the stalls people are finding. (Note: we idle at roughly 0.7 pounds per minute of airflow while WOT @ 7000 will be 34 pounds per minute for comparison).
Now the turbulence issue I mentioned before can rear its head. It is a combination of what air inlet hardware is on the car - K&N, a heavily oiled K&N (which coats the air meter element), or other intake system components.
Furthermore, stalling issues result from a stack up of errors. If you happen to get a high spec flow ISC valve then you will be less prone to stalling out with the screen removed. That is why some people here changed an ISC valve and haven't stalled. However that is expensive to try overall.
The air meter transfer function in the calibration is flowed with the entire air inlet system intact – from the silencer tube to the t-body. It is even positioned on the flow stand in the exact orientation as in the car. Changing the tube between the air meter and t-body can significantly affect the overall transfer function resulting in measurement errors. The basic rule of thumb is that any change within three air meter diameters (before and after the air meter) can have an impact on the transfer function.