Because you need to open throttle complete from specific load, not from pedal. It makes no sense to make boost with half closed throttle. Maybe open throttle from pedal much earlier and feed some external boost controller by pedal signal too, not from throttle position. But i guess it will drive like a very old turbo car with bad to handle boost, so hard to regulate torque.
You can feed external controller with pssol_w. This is what my boost pid code does.
Additionally there needs to be a patch to throttle controller so it knows what is minimum boost (gate boost) and force it to go WOT any time request is over that, otherwise it won't go WOT properly.
On top of all that, to calculate msdk properly it needs to know pressure pre-throttle and pressure in manifold.. If it knows base boost and base boost is not ridiculously high, then it is possible to do this fairly OK from a fixed map filled with the base boost pressure ratio, however the pressure in manifold requirement means that the model must be adjusted fully to show correct values OR it needs to be a speed density setup.
Let me put it this way - those of you who need to ask why this is necessary do not understand how the ECU works, and you're best case years away from getting it right, that is if you have embedded/comp-sci backtround and worst case light years if you don't even know asm.
Until you do, trying to do a turbo project on N/A ECU will end up in a car that drives like shit every time. There will be massive cuts on part throttle. It won't go WOT when needed etc. Or the throttle torque control will be utterly destroyed so it drives more or less like a cable throttle car. But of course forget automatic gearbox and ESP working properly.
Most of the cars you see done by various "tuners" drive like shit or have torque control completely destroyed.