Good work Peterae
The shorted winding is where the DC might circulate (should you get an induced DC current and get it circulating) and is possibly how SM gets his heating effect. I believe he may have picked off the DC component at 180 degree points of this winding. "Induction" of this sort, and I use the word very loosely, is not possible with normal transformer action as the result is always a null DC component once the system settles.
It is strange that you are getting such large output. The shorted adjacent bifilar wind should have clamped the output. Although your explanation of the winds is clear, it would be helpful if you could go back and annotate that original drawing with some text and arrows. I can see the horizontal outer winds. Are there any horizontal inner windings?
Try a very light load on the open ends of the large winding to see if will tame it a bit, probably 20k to 100k although you might not want to tame it, rather lightly couple it to the input windings. That will soak up and regenerate some of the energy.
As for the power dissipation in the 47 ohm, the temperature rise is a good rough indication of power and easy to compare to known heating from the same resistor hooked to your power supply. I would use a 1/4 watt if the heating is very small.
Thanks for sharing your work.
Just because it has a patent application or is patented does not always mean it really works.