Richard, You hit this nail on the head, for su

All the issues you focus on are quite particular and specific, and
perhaps too much so. I find them equally interesting and you ask
reasonable questions. Unfortunately, you question above suffers from
every problem imaginable for measuring, and any report you get without
considerable qualification is probably sheer fancy. It suffers from a
version of Heisenberg's problem of disturbing what you attempt to
measure, and invalidating everything in the process.

I don't have much trouble with the antennas I actually build. The
aluminum and wire I sling around tend to behave.

I do have some vested interest in overspecific modeling in the sense
that since I'm deploying all my antennas from the balcony of my
apartment (with its overhanging roof), and don't have much space to
build, adjustments and pruning are very hard to do. Some performance
gains can certainly be had by adjustment... but in my situation now
it's so much easier to push bits around in EZNEC.

This one, though, was just pure impractical theoretical curiosity.
That, and a little bit of being concerned that people seemed to "know"
the answer, that the answer was obvious, and, well, I didn't think it
was. The current in real experiment would be, as you pointed out,
pretty much impossible to measure without disturbing it.

But as I see it, from a practical standpoint, in a real system where
you're using the shield of a piece of coax as an antenna , the current
in the center conductor never matters.

The existence of the current matters, at least as a learning
experience. It matters in the sense that some seem to think that there
couldn't be a current, just because a piece of coax is "shielded". Or
because the center conductor is "floating". It's precisely the fact
that the center conductor is "floating" that makes the (probably) weak
coupling at the ends important.

Thanks for posting your results. What you describe certainly jives
with the fuzzy idea that I've got of what's doing the coupling ... or
at least where it's happening. Of course, the tube without center
conductor is certainly a waveguide well below cutoff, so we've got
solid footing saying a short wire inside a long conductive tube a small
fraction of a wavelength in diameter will have little current on it...
and the model also exhibits this.

I'm done with this one unless I find a leadless current meter (small,
battery powered with an A/D converter and optical fiber out?) and a
big piece of copper pipe laying around!

73,
Dan