G. White wrote:
"Output Z is irrelevant."

This is an old argument in this newsgroup. I became convinced long ago
that there are cases in which impedance is very important.

"Transmission Lines, Antennas, and Wave Guides" by King, Mimno, and Wing
make a clear and concise case for the princuple of conjugates in
impedance matching on page 43:

"If a dissipationless network is inserted between a constant-voltage
generator of impedance Zg, and a load of impedance ZR such that maximum
power is delivered to the load, at every pair of terminals the impedance
looking in opposite directions are conjugates of each other.

To secure maximum power output from a generator whose emf and internal
impedance are constant the load must have an impedance equal to the
conjugate of the generator`s internal impedance."

Radio transmitters don`t produce significant harmonics. It`s the law.
They are linear power sources. We can and do tune them for all the power
they will produce under their particular operating conditions of drive
and d-c power supply. They operate at more than 50% efficiency which
means that they don`t take power 100% of the time, but are switched-off
during part of the r-f cycle. Output impedance is thus an average over
the entire cycle. It`s OK. We have no harmonics. Gaps are filled by the
tank circuit and other filters. The radio is a proper source. The
impedance added by off-time is called "dissipationless resistance"
because no power is lost in the radio while it is switched-off.

Best regards, Richard Harrison, KB5WZI