In some ways, this is somwhat like the behavior of RF in a waveguide, or the tuning of a musical instrument. If the inherent resonances focus at a certain undesirable point. The output is altered in amplitude or frequency in any one of a number of undesirable ways.

Good analogy, except that the diameter of the pipe is not a factor if it's big enough. It is a lot like getting reflections back from a badly terminated waveguide -- and an exhaust system is completely unterminated, of course. In that case, length has everything to do with what frequencies are additive or subtractive, and therefore what RPM are affected positively or negatively.

In this case when the pulse or wave movement has enough force it can physically eject the fuel charge.

I suspect it's more critical in my motor than in others because we've tuned the intake length to produce positive pressure outside the air horns beginning at 6500 rpm or so -- sonically forced induction, in a sense. Not sure, but this might also mean that at 1/2 that rpm, we have positive pressure halfway down the intake length and/or negative pressure outside the air horns contributing to the reversion.

The header should be overcoming that, but something is defeating it.

Since you have, as I recall, a long straight section under your axle, can you use a telescoping section and tune it like trombone?

I like it! Unfortunately, the muffler is also behind the axle and there isn't room to move it fore or aft. The whole thing is going to have to end short of the axle. It's not a big deal to play with sections of pipe of various lengths, but it is a problem getting enough ground clearance with the muffler under the floor... the car is pretty low. There will be a way.