After numerous experiments with Ellis patterns in OrcaSlicer, I found that both acceleration and flow impact pressure advance on reciprocal curves. With enough data points I was able to easily and pretty closely fit a function using a linear solver (with 1/x and 1/y as inputs) to calculate pressure advance for any combination of speed and acceleration.

My question is if using a model like this for calculating pressure advance would be of any advantage.