I got an answer from my engineer friend, but havn't yet got his permission
to forward the entire answer.

His basic thinking is that we need to think in terms of fields, not current
flow and that fields flow outside but near conductors not in them.  Fields
will flow between conductors, E.g. in coax because this provides a low
impedance path.  This is hard stuff to follow, particularly in practical
terms.

He says, as we  should know, that ground systems don't work at frequencies
much above power line frequency and thus are of no use in keeping R.F. out of
where we don't want it.

It would seem that the quality of coax shield, and the location and
directivity of antennas
away from susceptable devices are the main methods of controlling field
flow.  Of course most of us can only do so much along these lines in the
real world.  

BTW, in testing cables, one thing we can do if we have an accessible power
meter and a dummy load is to run a little power through the cable in
question to a dummy  and see how much power we get back. This has been
discussed recently in QST. If a cable shows loss that power has to go
somewhere, either heating the coax or leaking into the free space around the
cable.

My friend Ralph, has the 6Th eddition of his book on "grounding and
shielding" comming out from Wiley in April.  When it's out, I'll request it
from Bookshare and see how much can be made of it.

Undoubtedly Bookshare won't make the diagrams accessible, but perhaps the
text will help.

tom Fowle WA6IVG