Can we draw analogy between em power flow through free space and ac power flow through a transmission line? Knowing that the free space has a characteristic impedance (which is purely resistive, measured in ohms) I was wondering if I can model the free space as an infinitely long transmission line- comprised of distributed inductors and capacitors.

We know, a lossless transmission line model of infinite length supplied by an AC source at one terminal looks:

If we consider the direction of changing electric field across the capacitance (say X axis) and that of changing magnetic field inside the inductance (say Y axis), we find the poynting vector directing outward of the plane of transmission line (Z axis)-- which seems weird, since power flow direction is certainly along Y axis.
However, if I adjust the orientation of the inductors in following way:

It seems that the problem regarding to power flow direction is solved apparently.
However, I am not still convinced with this depiction. Firstly, I have considered only the magnetic field inside the inductors. But H has a non zero curl, it ends upon itself. So if I take the total H field (around the inductor) into account, I end up with a zero power flow -- which is definitely not what is happening.
Moreover, the power flow, irrespective of the orientation of inductors, must be along transmission line (not outward of the plane containing TX line).
At this point, I wonder where I am making mistakes. (i.e.- is it because I am ignoring current through the capacitor and electric field across the inductor? )Or is it a crap idea to model the free space as transmission line ?  
 A: Yes, you can view a transmission line as a model of an EM wave propagating in one particular mode, in this case it is a plane wave moving in a given direction. The model can describe how the voltage and current of an antenna port would behave, especially, if you include several of these transmission lines connected to the port with ideal transformers of different turn ratios. One line may represent the main radiation lobe and is better matched carrying most of the input power, while the others correspond to the sidelobes. At the input port you may also place some other capacitive and inductive loads representing the so-called non-radiating near field.
A: I think you would very much enjoy Chapter 27 "Field Energy and Momentum" in Volume 2 of the Feynman Lectures on Physics.
What you are doing is precisely the kind of exploration that will build a deep understanding of the EM field. In many ways, it is a great model. 
You're running into problems simply because your transmission line is made of "lumped" elements, whereas the beautiful induction of electric fields by magnetic fields (described by Faraday's law) and contrariwise (described by Ampère's law) to give a self sustaining EM wave is not lumped. So for example, you have your magnetic fields contained mainly in the inductors and the electric fields in the capacitors. The freespace Maxwell equations don't separate the fields in this way. So, although your model will show the same transmission of energy as the freespace wave, you can't infer the field vector directions. 
You could try thinking the other way around: the freespace propagation mostly models a transmission line. With any transmission line, the wires - or conducting sheaths - only serve to set the boundary conditions which guide the waves which flow and are self sustaining anyway. In a gigantic, 500kV main trunk power transmission line, the energy from the power station is not flowing through the wires: the wires only let charges shift to set up the right field guiding boundary conditions. It's the freespace around the wires that is the conduit for the energy, which travels as a freespace wave.
