Wobbly, your reply #2850 seems to be just a construct to try to build a mental picture of the results of the recent dyno testing. However, I think there are some holes in the analysis and your thinking about what was going on. You might consider the following comments.
“The pressure waves travel down to the final expansion point and they are bounced back toward the intake valve at opposite polarity.” The pressure waves travel to any and all expansion points, the first and probably the most important of which is at the collector. This is why header length is critical to the tuning possible in the exhaust tract. The reflected negative wave starts back there, and should be timed to achieve the best effect. Generally, a megaphone immediately follows the collector which, by providing a gradual expansion of area tends to spread out the reflected wave in time, broadening the rev band over which the effect is felt.
“Gasses that go from a hot to a cold environment sort of "stack up" on each other as they he decelerate.”
Are we talking about gases or waves here? Presumably waves. In which case, they can’t overtake one another since at any point the temperature is the same and the wave velocity would be the same. They may slightly approach each other (in space), but how would the trailing wave overtake the leading one?
“The O2 sensor pipe was stuck more than a foot into the system and it indicated lots of fresh air intrusion with turbulence between pockets of it and hot exhaust gas.”
Postulating that atmospheric air has traveled upstream against a very robust exhaust stream strains my credulity. What physical evidence is there that this is happening? Was there a smoke generator at the exit that showed anything like this was going on?
It is far more likely that the O2 sensor is telling you that there is evidence of incomplete combustion, ie., residual oxygen in the exhaust, probably inlet charge lost during overlap. All of which brings into question the appropriateness of the tuned lengths and configurations of the various tested arrangements.
It is clear from the torque curves that the non-muffler configurations work at lower than desired revs and that they just are not right for other rev ranges. That the muffler works better may just be a happy coincidence of longer length working in conjunction with the header/collector length and some increased back-pressure having a serendipitous effect on inlet charge capture.