I had the thought; that if the feed point was made the source at the base of the radiator...then it would seem to me to just by-pass the trombone tuner. When I place the source at the base of the radiator I think I saw a better match than some tap points on the trombone, and you're right...the tuner had no effect when the source is not tapped to the trombone.
I used a split "Type" feature in Eznec...which allows the tap to get closer to a wires junction. Maybe your idea and this feature are similar. I don't understand the technical aspects of this part of modeling.
As noted above, this was my experience also...the antenna tuner had no effect on the tune with the source connected at the base of the radiator.
The radials on Jay's antenna attach to the side of the tubular mount. There is another similar model out there in the market that does have the radials symmetrical.
My model has currents turned on and I see very little current flow on the antenna view image, and the tabular currents data shows very little currents too. I'm of the opinion this all indicates the ideas when CMC manifest or not, but that is just a guess.
.58wl is pretty short, but I have an idea why this happens. I suspect that a model that shows the worst match that the model likely will show a little increased gain compared to another similar model that has less impedance transformation requirements at the feed point.
However, when I test the idea by skewing the antenna construction a bit to make the match worse...it will change the match, but the performance results often stays the same or makes only a really small difference.
Just a few days ago I made a V58 model for Needle Bender regarding his issue on the forum about 5/8 waves, the Maco V58, and his idea about .64 wave, and I posted the results. I did not see any added benefit to the Maco, but I hear reports that it is a really good working antenna.
After I did that project...I was curious about adding the matcher to the V58 and I was pondering the same thing as you...where is the physical bottom of the antenna, how far up is the tap point on the radiator, and how do I account for the area that allows the up and down adjustment of the radiator that adjust the reactance...as you say the capacitance area. I'm still stumped or else I have a mental block on this one.
I'm also not sure how the different mix of tubing and flat bar stock in the feed point area will effect results. I have a feeling, working with the trombone, every little material detail makes a detectable difference for the model.
IMO the hubs on CB antennas we use likely have more important effects than I imagined in the beginning of my modeling. I think all the little parts around a feed point with high currents flowing...can make very sensitive responses.
