Richard Clark wrote:
. . .
On the face of what is offered at Tom's page, there is an immediate
and irrevocable error of 5% built into the instrument as described
sitting in its calibration fixture. Under other circumstances, that
error could easily eclipse 100%. Suffice it to say it will never
achieve better without a small book of charts.

As I offered, accuracy is often claimed, but rarely (never) proven.

This is a simple counter-proof. Tom expresses it without being aware
of the implications:
T1 is a current transformer. ...
When the single turn primary (a whip or mast) has 1 ampere,
the secondary will have .05 amperes (inverse of the turns ratio).

This means the turns ratio is 1:20.

all very standard stuff as you may well note. Going on:
This type of meter is much more reliable and linear
than thermocouple RF ammeters, and perturbs systems much less.
This, of course, is related to the "stated but not proven" class of
statements that litter the WWW (much less this thread).

Here is the literal error:
I've applied 50 watts to a precision 50 ohm load,
making wire current 1-ampere.

Let's assemble these statements. We have a current transformer. It
is loaded with 100 Ohms with a lightly coupled linear indicator. It
has 50 Watts applied through it to a load. That load is 50 Ohms.

What the meter should indicate is a current of 0.9535A if we are to
believe that the 50W is absolutely accurate (it is not, but we will
skip that for another discussion). The author, Tom, offers to trim
the potentiometer for a 1.000A reading - WRONG!

How can this be? It is all in the statements offered above. The
current transformer is also a RESISTANCE TRANSFORMER. That 100 Ohm
load to its secondary is cast into the primary as an in-series 5 Ohm
resistor adding to the 50 Ohm nominal load.

The resistance is transformed in the ratio of N^2:1, which is 400:1 for
a 20:1 turns ratio. So the insertion resistance is 100/400 = 0.25 ohm,
not 5. I'm confident it works as Tom claims.

. . .

If one were to claim to have made ANY current measurements, and then
wholly ignore the contribution of errors, then the discussion of phase
in a system such as a 5 Ohm radiator with a 5 Ohm instrument loss is
going to be absurd. Such are the fruits of poor reporting.

No, this is the fruit of poor reporting.

Roy Lewallen, W7EL