Frank Berto did something amazing. He took thousands of measurements. It would be multiple adjustments of tire pressure to obtain each measurement. He did it on his own time and on his own dime. That the result is straight line graphs demonstrates he did his work well. Straight line graphs are also a rather elegant demonstration that simple physical forces are all that is involved here. Basically it is gravity and the gas laws applied to inflating bicycle tires. If you want to argue with gravity have at it.

Of course a large data set has a few oddities. Berto measured the inflated diameter of tires. The governing assumption was that tires are approximately cylinders. Tire width was a proxy for the volume of air contained in the tire. That is not going to be perfect. Tires are not perfect cylinders. Some tires are tall, some tires are squat. Rim width, bead design, hook shape make a difference. Clincher rims have a well, the well has air in it. Any test protocol that accounted for all that would be complex and ridiculous. Tire width was measured because it was possible and it was reasonable.

The commentariat here is busy focussing on third order variables. And on the magic eightball. To anyone who thinks that weight does not transfer forward on braking, or rearward uphill, or that tire sidewalls by themselves support weight, or that they will change tire wear rates with minor changes in pressure, or that a 5psi differential between front and rear tires does anything at all, to those people I have nothing to say. The magic eightball is your friend. Your ouija board will give you good numbers for inflating your tires.