My main hang up was wrapping my head around how to design the mapping table, but I may have just figured it out:

So baseline is 1 psi governor pressure per 1 mph (60mph = 60psi).

At higher loads, we want a lower governor pressure to raise the shift points, so something like 0.8 psi/1mph (60mph = 48psi). At even higher loads, we may want even lower governor pressure, so 0.6psi/1mph. So if you floor it from a stop, the governor pressure will rise slower than if you drove off normally, and with the pressure being lower the shift points will be higher.

So, a chart with a multiplier as the y-axis, and Load (MAP or TPS) as the x-axis.

The y-axis is centered around 1, with the curve able to go to lower (or higher) multipliers at higher load values.

So, I look up the current multiplier for the current load, and multiply that by the current speed to get a target governor pressure. Then I command the governor solenoid with feedback to reach that target pressure.

At lower loads the multiplier can be adjusted up or down (greater than 1, or lower than 1) to get the desired shift points.

Questions:
1. How is downshifting from 3->2, or 2->1 done internally when under load? Does it happen at all? If you're puttering up a hill in 3rd, then floor it, does it downshift to 2nd, or to 1st?

2. Should I have the option of forcing the governor pressure high to force an upshift at a certain (high) RPM?

3. Since lower governor pressure leads to higher shift points, what would happen if I commanded zero (or very low) governor pressure throughout the MPH range?