This is likely what Benny Parsons and the team at Edelbrock saw back in the '70s when they
Reversion Effects From a Restrictive Exhaust System
Generally speaking, as backpressure is allowed to increase (largely based on header primary pipe size), cylinder pressure at the point of intake valve opening will correspondingly increase the energy level of reversion pulses delivered back into the inlet track. There is a balance that needs to be struck between primary pipe size to effectively improve the exhaust "blow down" period and proper cylinder evacuation. This becomes particularly critical during mid-rpm operation where over-sized headers provide insufficient flow rates for good cylinder cleansing of exhaust gas. Think about the time when smaller primary pipe size improved mid-rpm power with little or no attending loss at higher engine speeds. Reports of such observations among engine dyno operators are quite common.
Reversion Effects From Improper Valve Timing
This is probably a good place to reference the accompanying sketch (Figures A and B) showing a simplified intake path pressure trace from intake opening to intake closing as a function of rpm (time or crank position). Note that there is an initial spike that represents a reversion pulse entering the inlet track.
Until the point (rpm, time, or crank angle) is reached when this pressure pulse decreases to equal that of atmospheric pressure, there is no flow toward the cylinder. But from then until intake closing, pressure in the intake path is less than atmospheric, allowing the cylinder to fill in a normal fashion. But you'll also note there is a second "bump" in the pressure track at the point of intake closing. This so-called "hammer effect" is similar to what you hear when quickly turning off a water faucet. Sometimes, this pulse is considered to be the source for reversion, but it is not, although the wave it creates in the intake path will oscillate back and forth in that passage until the next intake opening. To further pursue this would lead us into discussion involving Wave Motion tuning and not the purpose of this story.
However, using this same pressure trace example, you can begin to see how the intake opening point could affect reversion pressure. For example, the earlier the timing (either from cam phasing in the engine or specific valve events), the higher the potential cylinder pressure at intake opening. Designing or modifying intake valve seats and heads to decrease reverse flow at low valve lifts is an effective way to help dampen reversion pressure and allow for advanced cams or early lift points to be less problematic. Connecting rod length and overall rod/stroke ratios can play a role as well. It's usually a good idea to have some of these types of discussions with your cam supplier of choice. The more knowledgeable ones will have some specific comments to share on the subject.
How Reversion Becomes Contamination As a Function of RPM
We'll leave you with a few thoughts about this issue, but it's at the core of how reversion should be viewed and handled. Normally-aspirated engines are particular culprits for the conditions of reversion. Even though there is a proportionate increase in the number of reversion pulses as rpm is increased, there is less time for these pulses to re-enter the intake track. The dominate variable in this set of conditions is time. Stated another way, the degree of encroachment to which reversion pulses can go back in a counter-flow direction is shortened with increased rpm.
In some early reversion studies in which I participated, we were using an inverted, transparent, bowl-shaped cover above an intake manifold's plenum area. With the engine idling and camshaft somewhat advanced, you could actually see little vapor "puffs" blowing back into the plenum area, obviously following the engine's firing order. As rpm was slowly increased, the extent of plenum penetration by these puffs grew less and less until they were no longer visible. Further increase in engine speed caused these reversion pulses to diminish into merely becoming contained as contaminants within the combustion space where they remained at higher rpm. The conclusion we drew was that the effects of reversion never completely disappeared, becoming contamination for fresh air/fuel charges throughout the rpm span.
So, reversion is a problem. It consists of combustion byproducts that can reduce net power. It needs to be identified, recognized, and dealt with by whatever means that will cause it to be the least effective. It has specific and tell-tale signs, some of which were pointed out in this story. But the overriding point is that it is a subject not to be ignored in the building and tuning of a race engine. Its effects can be minimized, once they are understood.