My understanding of the turbo upgrades is that the three stages are "sizes" of the turbo (rather, turbine / compressor pairing) such that the peak efficiencies (hence torque) sit in a different part of the rev range depending on the kit you buy.
This is why the kits are low / mid / high rpm, since they are tuned for volume throughput, which ultimately only determines their
maximum possible boost on a given-sized engine. Tweaking the wastegate (i.e. how much of the exhaust throughput actually gets used) allows for boost control - i.e. the "bump" in the torque curve produced by the action of forced induction can be scaled so that it blends into the (pseudo-) NA torque curve (no boost), or stands out on its own (maximum boost). As a result, the boost controller makes more sense now than it ever did in any previous version of GT, where turbo upgrades were sold in "stages", which could mean anything.
Obviously, there comes a point where scaling the exhaust flow via the wastegate leaves too little volume to actually push the turbine around and compress the charge air, hence dropping boost off dramatically. That is, the minimum practical boost is also determined by the kit size, and should be lower for the lower rpm range kit than it is for the high-rpm kit. Ideally there would be an overlap of practical boost ranges for each kit, meaning that the same peak torque figure should be possible on two adjacent kits, but the location of that peak in the rev range will be different.
In short, I agree that boost control is very relevant and useful in tuning cars for many reasons, not just peak numbers. I don't like the way the mid-rpm kit tails off at higher rpms, for instance. I'd probably prefer to run the high-rpm kit at a lower boost level - I know I could use the power limiter, but a wastegate actually acts as a
torque limiter, so it's not the same. Then there's the possibility of
welding the wastegates shut, as they did in F1 and LeMans in the '80s (in particular), and some of the Group-C numbers only make sense with that in mind. It should be noted that the hardware didn't like this maltreatment, so I agree that engine (and drivetrain?) wear should account for this somehow.
The torque-scaling resulting from boost control in GT seemed sensible and "realistic", so I don't see why it's been excluded ever since; except for the engine wear issue.