It is not really necessary if you use the same cabinets and drivers. I measured the drivers in my backyard, on axis and off axis, and used those measurements to help me choose the appropriate crossover frequencies for my drivers.I understand why one would model the box, drivers and crossover when designing a passive speaker from zero. But if theres is almost any crossover available in the dsp amp what would still be the purpose of modelling? And if, as in my case,I have the drivers and the box as a given, and any crossover available in dsp, what and why would I still model then?
All of my final tuning I did in my listening room at my listening positions. I have one tuning specifically for serious listening with the speakers pulled 1m into the room and me in my listening chair, and another tuning with the speakers near the wall and adjusted to give a good average response over a larger area.