Well, I have no measurements that disprove your thesis, and as I've pointed out, what's going on in there is obviously (now that I have one to try and measure on) far more complex than any armchair theories I've seen here or at fusor.net, certainly.
I do however see decent evidence that quite a lot of the ions are produced partway down the accelerating well -- many right at the bottom. I don't see any evidence of any bunching into packets with broadband antennas and fast detectors (up to ghz resolution, and one of my gigs was tempest antenna design for electronic warfare). I looked for any periodic recirculation really hard -- and can't find it. I see no evidence of any bunching (other than into the rays) whatever, and I've been looking hard for it -- and even trying to create it with active drives -- which, when successful, vastly increases the Q I see - but only for as long as I drive the bunching, no after-ringing whatever. In fact, there's no mechanism I'm aware of that would produce it without external drive, but plenty of things to prevent it happening or staying if it is made to happen some way. Any charge separation in a plasma tends to self destruct pretty fast unless continually driven. I do measure extreme electron excess no matter where I put a probe in a fusor, even though they are continually driven to the walls. Most seem to be coming from hits on the grid, which has a nicely infinite supply of them. Grids that get less hits by whatever mechanism (for example, having less area to hit or giving better focus) reduce this measurably.
Whether you're right or not -- the sure thing is that it's so far so complex in there that no simple feed-forward explanation seems adequate, and we'll only do the old head-slap in hindsight after getting a lot more data than anyone yet has (that I know of). It could be that simple, dunno. Doubt it, but there's no real evidence either way at this point that I've seen. It's like trying to look at a starting pattern in Conway's game of "life" and predict what's going to evolve -- emergent behavior is always tricky in foresight.
After all, it only takes 6-7 very simple equations in few variables per cell to predict the weather accurately - but you'd never derive them by watching clouds and guessing. Even if you know them, I'm sure all are aware that you need to be able to do quite a lot of them very fast to simulate weather faster than the system itself does -- for a lot of things, the universe is its own best simulation, and it automatically runs real time.
This seems like a very similar problem -- but just working out the effects on one charge of all the others is a lot harder than the weather situation (eg things don't divide nicely into cell interiors that don't interact with other cells except at the edges), and that's only one of the equations you need. Looks like about the same problem here -- just different math, harder to do because of that non ability to divide the grid up and work only within one cell at a time. Or so the gents at SIMION said as well.
Which is the point of this direction change here. Rather than hoping that whatever equilibrium a fusor obtains will magically create whatever the best fusion conditions are seems like it's reached whatever limit there is on that without a lot more understanding, and probably some form of active input, rather than just letting it sit at equilibrium. In fact, quite a few experiments here show that practically anything that kicks it off the stable spot improves things till it all goes back - which happens quite quickly. Doesn't seem to matter much at this point what kicked it off -- magnetic pulse, field perturbation, whatever, even DC magnet with some weird orientation seems to help. It's as though the "stable" state of a fusor is actually the very worst case for fusion, though it acts like what the math guys call an "attractor" and the fusor "wants" to be in that state unless you kick it off of that.
Whether that's due to your proposed mechanism, some other one we haven't thought of yet, or the spins/selection rules kinds of things -- no clue at this time, I don't think anyone does have one. Hence the quest to find that out in the simpler system where I have a shot at it.