Antonio,

The carbon fiber rod is very thin with respect to a wavelength. There
won't be any significant of the RF current variation around the
circumference of the rod.

If you have the wire insulated from, but closely spaced to the carbon
rod. Then the rod becomes a parasitic element ONLY coupled to the wire
via the fields. If the rod is insulated from ground by a base
insulator there is probably no problem. I think the current will still
divide approximately the same way. There may be some subtle
differences.

The situation where you allow the rod to touch the ground (or worse,
the ground radial system) is drastically different. Now you have,
instead of a monopole, a weird parallel transmission line. A closed
top gives a folded monopole, but opening the top (insulating the wire
from the pole everywhere) doesn't change the fact that it's a
transmission line.

Imagine the following: the top is open, you're operating on 30m where
the pole and wire are very close to 1/4 wavelength long. This gives a
very *low* impedance at the bottom of your open transmission line stub.
I think a great deal of differential mode current could flow in this
case, and I think it would cause a good amount of loss.

I think your plan of having the wire connected to the pole at top and
bottom is sufficient if you have a good base insulator.

Dan

P.S. I wondered about the surface of the rod because I thought perhaps
there would be an outer nonconductive layer of thin plastic film to
protect from carbon fiber splinters. If there's no such film, then
spiralling the wire around the pole or clamping it periodically to the
pole will make for good contact everywhere.