If your superconductors are truly lossless*, then any radiation resistance will be a considerable fraction (that fraction being 1/1 :-) ) of the total. So you've got it right. :)
*Real superconductors aren't, not at AC. This is macroscopically visible with type II superconductors, which exhibit flux pinning (a hysteretic loss effect that goes down to DC). At AC, the resistance equivalent is nonzero. (Apparently the loss at DC isn't exactly exactly zero either, but it's certainly good enough to leave NMR magnets charged for a long time.)Type I superconductors are generally quite good, at least at liquid helium temperatures: resonators for LINACs, in the 200-1000MHz range, typically attain a Q factor in the 10^7 range. But it's still finite. A true superconductor shouldn't depend on anything, but apparently they achieve a couple times better performance (to say 3 x 10^7) by polishing it, but not too well. Or something. I forget exactly, but surface finish is involved in any case. Chemical purity matters, too.
This is all very poorly understood physics, so I'm afraid I can't give any better underlying answers. But the effects are real.
Yeah, if you have superconductors, you can dump insane amounts of reactive power into a loop, and assuming your amplifier recycles that reactive power rather than wasting it... then real power is still real power, and it can only be due to radiation resistance, however small (almost).
Tim