As the incidence increases, the airflow circulation provides an increasing pressure reduction above the forward LE area of the upper wing surface. This accounts for the pre-stall movement forward. You can see this in any of the typical pressure plots, eg,
www.google.com/search?q=wing+pressure+pl...imgrc=BhJZciXzGZJiAM: this URL provides a number of illustrative diagrams. The URL cited below also gives you an idea of the variation with incidence.
The downside of increasing the wing incidence, though, is that the airflow aft of the LE has to contend with an increasingly adverse pressure gradient which is trying to stop the airflow's aftwards movement. Eventually, it all gets a bit too much for the air (at stall) and the aftwards movement of the airflow near the upper wing surface slows to the point where it stops and separates from the wing surface region. The end result is that the low pressure region towards the LE collapses and the pressure plot changes significantly and fairly rapidly Unfortunately, I can't find a suitable pressure plot sequence (plenty of line plots but they aren't as intuitive) so imagine that the upper forward reduced pressure area reduces suddenly and becomes a bit like the low incidence plots shown here
avstop.com/ac/flighttrainghandbook/pressuredistribution.html. This aftwards movement of the pressure plot is associated with the aftwards movement of CP and a (very desirable) nose down pitching moment which helps resolve the stall problem.
To get a good feel for what is going on, you really need to see the variation with incidence in a wind tunnel setup. I have seen videos of this on the net but am not having any luck finding any at the moment for you.
