Cecil Moore wrote:

I suspect that all necessary reflections take place
at the impedance discontinuity and the characteristic impedance
is established immediately.

Cecil - you might like this.

What I got out of Feynman's QED was that the probability for a
reflection to take place would be highest at the discontinuity. But
that reflection would take place at any point thereafter with some
lesser probability - inverse with distance (smaller arrow, different
angle). Summing all the little probability arrows (vectors) for various
distances results in a magnitude equivalent to a reflection occurring
solely at the discontinuity. He used the reflection from a pane of
glass as an example. But as you know a change of index of refraction is
directly analogous to a change in impedance.

In relation to this discussion, there would be some distance from the
discontinuity where the probability of reflection is so low, that beyond
that point, the sum total contribution of all the reflections is
immeasurably small. I believe it may have been Reg Edwards (sk) who
spoke of the distances you refer to. The brilliant thing is that Quantum
Electrodynamics substantiates his argument.

73, ac6xg