I think you are getting a good grasp of the issues.
I always thought that curiosity was one of the best tools to gain knowledge.
What you are asking is difficult to answer easily, succinctly, and quickly.
Perhaps you can build on this:
In an air stream where air as been made turbulent, the turbulence extends beyond the obstruction that caused it (in this case the filter media) at a length determined by the velocity of the air. Within duct work, it is also related to the volume or mass quantity being moved.
At idle where low volumes of air are moving, the turbulence may extend only a short distance from the filter media. At high RPM, where the engine is moving larger volumes of air, the speed through a fixed size duct greatly increases, and the "fingers" (if you will) of turbulence from the filter or other obstruction/feature causing impedance, extend a greater distance.
I haven't done personal testing, but the stock arrangement may have an RPM point where the plenum isn't large enough to prevent turbulence from reaching the carb throats. Presumably, that RPM/air volume/speed is above what occurs up to the stock red line with a standard displacement engine and cam. I'm assuming the teams of Honda engineers and test technicians did the right thing for their machine. But, if you modify the engine volumetrics, you may well need something different regarding induction design. There is no one device that does all things well in all situations/conditions.
Hope this helps.
Sorry, I have no web site or book. I'm thinking about it, though.
Perhaps you can identify what mods or function you have that require more air flow than the stock air induction?
Cheers,
