With modern aftermarket bits, it seems entirely possible to take the 89 mm stroke Commando engine to much higher rpms than even the factory race bikes of the period (and serious privateers like Ron Wood, etc,). The real question is are they making any more horsepower at higher speeds (like 8500 rpm). Other than the occasional need to over-rev going into a corner, to avoid changing gear, or going a bit past the power peak to stay in the power band after shifting, what is the point of revving farther past the horsepower peak? I think it is going to be pretty difficult to come up with a build for an 89 mm stroke Commando that has a horsepower peak much above the 7000 - 7500 rpm range, and still have a rideable bike. Maybe for a drag racer or landspeed bike, and with a close ratio gearbox, but not for a street bike or road racer, where you need a broader power band. If you have a head that flows enough to make peak horsepower at 8000+ rpm, it's probably not going to have enough flow velocity at lower speeds to run well. There's also the limiting effect of ring friction losses at higher rpm, and we've covered that pretty well in other threads. Again, this is with 89 mm stroke. With the shorter and shorter stroke Commando engines we are building now, it's a whole different story.
It looks to me like the principal advantage of the XR design is that it allows a larger short side radius going into the valve seat. It's not really a matter of getting enough cross-sectional area in the port. You can already get that by opening up the existing ports, but it's difficult to do so and not get flow separation and turbulence over the short side radius.
In any case, like Jim, I'd love to see someone do all the work it would take to try to apply the XR750 port design to a Commando head. It would take some serious flow bench and dyno time to sort it out, but it might work really well on a race bike. Then again, it might not, but that's why they call it development.
Ken