Empirical trial may be the only means. Resistance checks of several concentrations of the salt might be scalable. The resistor likely has more than two electrodes.
Simply, the rate of change of fluid temperature versus acceleration time seems like the biggest unknown. Six gallons of a salt solution, 559A 3ø, time to acceleration of a ‘large ring mill crusher’ and resistor energy dissipation (a/k/a fluid state change) may not result in stable operation, making it infeasible. Sorry, but defining anything for “area under the curve” integration is out of my league. I’d take advantage of increased physical isolation during starting.
The span between 1.5x inrush and stalling could be narrow (or negative; ie, not possible.)