Wow, great info! What I could follow of it, anyway. It makes sense-because a running engine has it's intake and exhaust flow as periodic events rather than continuous, the peak must necessarily be higher than the average. That being the case, designing stuff only for the average would not seem to be the best approach. I've wondered about this stuff off and on for years, and my basic theory is that (being very general and vague here) stuff like what you're talking about may help explain why chrysler engines seem to usually do better than ford and chevy under ACTUAL operating conditions. Perfect example-last night I'm reading a book about how to hotrod smallblock chryslers, and the author mentions something I hadn't paid attention to; the fact that an LA engine's valves open "on center" with the cylinder bore. Meaning, the valves aren't crammed toward the intake manifold OR the exhaust manifold side of the cylinder (as viewed from above). This shrouds them the least possible, which ford and chevy apparently do not. So, with a given bore size and valve size, an LA motor shrouds the valves less. The result-more flow UNDER ACTUAL OPERATING conditions. This is guaranteed to be missed by a flowbench, since they don't put a cylinder against the head in question.

Now that it seems like I rambled, let me try to tie it all together; What I just said has in common with your post the fact that there are things our simplistic models leave out. Flowbenches don't consider the relative centering (or lack thereof) of valves in a bore, and many theories and formulas neglect to consider that gas movement in an engine is a series of pulses instead of a continuous flow. It is this type of oversimplification that causes us to have an incomplete and flawed understanding of what's really going on in an engine. Hence, a well-prepped 340 will usually outrun an equally well-prepped 350 chevy, and many people will wonder why. Well, THIS IS WHY.