[gone-ot]

...with "headers" there are two movements going on: (a) an acoustic 'wavefront' that moves through the gas at about 1700 fps (speed of sound in hot air) and (b) the mass movement of the gas slug itself as it moves down the pipe at about 300 fps.

...pipe diameter affects speed of mass movement down the pipes, while piple length affects acoustic "reflection" (Helmholtz tuning) time and strength (camshaft valve timing comes into play here too).

...too large diameter pipe and the mass movement slows down below the optimum value of about 300 fps; too small diameter pipe and the mass movement increases above the optimum value and 'back pressure' builds. 300 fps seems to produce the best "scavanging" action between the exhaust gas flow and the cylinders.