What's the difference between "in reality" and "the energy flows in the copper"?
In that the properties of the dielectric (insulator) and the geometry of the coaxial cable affect the energy transfer capabilities of the coaxial cable more than the core conductor does.
Really? If you replace the core by some plastic, what happens?
It becomes a wave guide. It is only practical at microwave ranges, because in the absence of a conductive core, the dielectric absorbs the energy in the EM field at most frequencies; and the characteristics of the outer "ground" become absolutely crucial, as full reflectance is needed for a viable waveguide. (We are talking about alternating currents here, after all; since as I already wrote earlier, for direct current, the energy does indeed flow in the conductor.)
More interesting is to examine what happens when you have flaws, say a short break, in the core conductor. If "the energy flows in the copper", then even a micrometer wide break in the core conductor would stop the energy flow, wouldn't it? It doesn't (for AC; it would if this was steady-state DC). It does cause all sorts of reflections and whatnot in AC, but a large fraction of the energy still flows.
This is an excellent example of why it is the geometry and not just the conductor that matters. It is silly to even attempt to say where the energy flows, unless we exactly specify the geometry of the system we're talking about, since it really does vary from system to system. The movement of charge carriers, current, is just the easiest way we can exploit the energy flow, make it do useful work; but it isn't exactly how the energy is always transferred within the system. Sometimes it is, sometimes it isn't: it depends exactly on the geometry/setup of the system.
Naej, I can't tell if you're agreeing with me and just directing the discussion using the Socratic method, asking genuine questions, or whether you have observed a flaw in my reasoning but are unwilling to point it out. Would you mind telling me?