[oldbeaver]

Dick,

Can you explain a Little more your idea?

You wrote:

"The ECU output that modifies the fuel flow does so by moving the rack so no need to measure ECU also. Rack position and the calibration curve for the pump will give fuel per injection stroke."

There is a wire between ECU and RPS. You can intercept signal in any point of this wire and it is the same.

I agree that calibration curve of any single injector of the pump are about the same (should be identical). The sum of all them is the curve of the pump. This curve is about quadratic, concave, with máximum around 2500 rpm.
But this is true for a singular load. A havier load will shift the curve upwards.

I have an idea for the math expression to model it. But you need a programable digital sampler to make the calculations to generate one unique variable meaning liters/time. And a variable that informs of throttle position as an indicator of load.

Any ideas?

Oldbeaver