"A related cause for unexpectedly poor fuel economy involves an inherent
weakness of the Wankel rotor design when used with conventional fuels.
Some studies have indicated that at high speeds, the rate at which the
volume of the combustion chamber increases in the moments after ignition
actually outpaces the expansion of the burning fuel. The result is that,
at high speeds, less useful energy is extracted from the same volume of
fuel, as the exhaust has to expend time and energy "catching up" to the
rotor before it can accomplish any work."

If that is the case then why is the HP of this 12A p-port engine
continuing to rise in almost a linear fashion all the way to 8800 RPM?
I think this is BS.

Paul Lamar ...No rotor no motor.

I had already removed it long before I got your post. My reasoning was
that all engines lose power as you get away from their peak on the
power curve and therefore not something new to the Wankel. That's the
definition of peak. Feel free to change anything on that page you
don't like. You are the expert.
-Jeff (Jeff Martin)

I tried that once and it did not seem to take.
There are several things going on here. Intake organ pipe tuning
is one. Limited air intake area as a ratio of piston area is another.
This is mainly a piston engine problem. Volumetric efficiency problems
such as forcing air to flow around obstructions such as right angle
side ports in rotaries and intake valves in piston engines.
P-port engines with large diameter runner/ports do not seem to suffer
from VE limitations. The VE/torque peak seems to happen at the tuned
length of the pipe over a wide range. In other words shorten the pipe
and the torque peak will move up generating more and more HP at higher
and higher RPMs until you run into rotor bearing load limitations.

Friction HP losses increases with RPM. Even if the friction torque
is constant HP =  Torque X RPM / 5252.

I think the Ingersoll Rand engine was limited by detonation and
not flame travel problems. I could be wrong. Also it could have been
limited in RPM by the rotor bearing loads.