This discussion about ideal sources and power absorption would be rather
amazing if I hadn't seen so much of the long sad history of this whole
thread. It's just another diversion to deflect the discussion away from
some of the sticky problems with alternative theories.

It looks like Cecil could benefit from momentarily abandoning his power
waves, virtual reflections, photons, s parameters, and constructive
interference, and go back to basic first or second semester electric
circuit theory. Connect an ideal voltage source to a series RL or RC --
all elements lumped, not distributed. Find the power P(t) at each of the
three nodes. (It's easy. Just calculate v(t) and i(t) and multiply the
two.) Remember that the sign of P(t) shows the direction of energy flow.
If I had the patience and self-control to participate in this endless
thread (and I admittedly have neither), I'd refuse to say another word
about it until Cecil shows that he knows what the power waveforms are at
those three nodes. Why argue with someone about distributed networks who
hasn't mastered lumped circuit analysis?

Roy Lewallen, W7EL