Yeah, I guess the only way to quantify a minimum-pressure is to test it. Maybe spin a wheel on a drum with vertical loading that's representative of typical users. The drum can have rocks glued to it and potholes dug into it. Then spin the drum at various speeds and run over the rocks and potholes. Do braking-tests from various speeds as well to come up with some sort of tyre-creep vs. braking-force vs. pressure 3D chart.

Then figure out some minimum pressure where the negative effects fall into acceptable 80% percentile rates?