...A true Windom, or OCFD (same thing) does not have a vertical radiator...
You're not going back far enough in time.
The original Windom (as described by, but not invented by, Loren Windom 8GZ(?),
back in the late 1920's or so I think) was a half wave antenna with a single wire
feed from the output of the transmitter to the magic tapping point on the antenna.
With the right tap point, the standing waves on the vertical wire were minimized,
so that part radiated more as a traveling wave antenna.
This worked reasonably well on 40m and 80m, where the single wire feeder wasn't
too long (in wavelengths), and often could be matched on multiple bands because
transmitter output circuits had lots of tuning range and hams only cared if the
transmitter could produce rated power - SWR meters were rare. In Europe it was
popular to feed it with a wire that was half the diameter of the antenna wire, which
was supposed to improve the match. Of course, on the other bands the whole thing
worked mostly like a random wire, sometimes with most of the radiation from the
vertical wire, but it did radiate. Coax cable was not as common for hams prior to
WWII, and this was often a reasonable approach that eliminated the need for
building open wire line for a doublet.
The idea of feeding it with twinlead as an OCFD seems to have popped up in the
1950's or 1960's. It was shown in my 1968, ARRL Handbook
, and was the
first antenna I tried as a Novice since my transmitter had a parallel-tuned output.
I never got it to work.
But that was also prior to effective current balun designs. At some point, perhaps
in the 60's or 70's, we saw the appearance of the coax-fed OCFD design with a 4 : 1
voltage balun at the feedpoint, wound on a ferrite rod. (Toroid cores were still
uncommon.) Because of the unbalanced load, this would have had a lot of common
mode current on the outside of the coax.
I really don't remember OCFD designs becoming popular until at least the 1980's
as more hams were using solid state rigs.
But the point of that tangent down memory lane was to point out that the "Windom",
though bearing a superficial resemblence to an OCFD, used a different feed system
and, while operation on the fundamental band might be similar, on other bands the
current distribution on the two wires was often different, resulting in different patterns.
The original Windom certainly wouldn't have had a low SWR relative to 50 ohms over
multiple bands - the feedpoint impedance was closer to 600 ohms, fed against ground.