Not so hard, really; imagine an X/Y graph with effective spring rate on the vertical axis and wheel travel on the horizontal. On "F" you have a relatively straight line rising to the right, whereas on "P" you have a line that curves upward, starting out flatter and lower than the F line, but curving upward and getting steeper so that it crosses the "F" line somewhere in the middle of the travel range, indicating a stiffer effective spring rate. That's why they call one
Flat and the other
Progressive.
"P" uses the linkage to change the effective motion ratio as the swingarm moves upwards to make the rear end stiff enough to support a passenger while making it soft enough for road use by a solo rider. What physically happens is that in the early part of the travel, 10mm of axle travel moves the shock less than 10mm of axle travel in the latter part of the travel does. In the end, you have the same amount of total wheel travel and shock travel no matter which position you put the link in, so the shock travel you don't use in the soft part of the curve you give back in the stiff part. Voila!
F does start out firmer as it's intended for track use, where you're putting a LOT more force into the suspension and want that in order to stabilize the chassis, but you get more consistent action throughout the travel with it; you just need the right spring for your weight & use. P is intended as a compromise for multipurpose use, and since most find the bike oversprung out of the box, they notice the softer initial action and think it's better.
Most sport bikes on the market come with progressive links on them, and most suspension shops sell flat (or flatter) links to make them work better for track/solo use. That's not by accident...