Making some assumptions about a sensible implementation for the test, an
interesting test Tom, thanks for the writeup.

One of the interesting propostions emerging in the discussion is that Zs
may be dependent on drive level. Your test setup would be an interesting
one to explore that effect (changing drive level alone, no other adjustment
or change).

In my own tests looking for Pf constant (independent of load changes), I
have noted that I get different results from the IC7000 at different drive
levels. Not surprising, as the gain of the active devices are likely to
change significantly in the upper half of the rated power range, and the
effects of either current or voltage saturation are likely to be manifest
at near rated output, and power control loops at maximum output.

If Zs is dependent on drive level in a significant way, then one must ask
what is the application of Zs in the case of an SSB telephony transmitter.

Owen