I haven't had time to fully read the replies to my post or examine Michael's revised figure, but I did want to add two more things.
First, I had some doubts about Shurcliff's numbers, so I decided to make some measurements on one of my cars (Ford Escort Wagon):
tire pressure = P = 32 psi (all 4 tires)
width of tire = t = 6.5"
diameter of tire = d = 22"
So that gives not 2000 lbs but rather 32*6.5*22 = 4580 lbs for air pressure force on half of the tire. Shurcliff's value seems *way* too low for any real car.
Second, I still have an important question in my mind that I don't think has been answered: If the tire pressure (P) times the contact area (A) of the tire with the ground is *not* equal to the weight (w) of the car divided by 4, then what *does* determine A? In particular, what do you need to know to predict it? For comparison to my car above, note that:
w = 3480 lbs
and I assume A = t*L where I used a ruler to estimate that L happens to be roughly equal to t for my car. (This value of A is probably an overestimate, because I haven't corrected for the treads.) -Carl