Hi George,
I have done much in asking Mike and Remko how this POTY works, but never got a good answer.
Than i took a more practical approach.
I have test build over 10 pieces for just finding what works and what does not.
I do not have expensive antenna simulation software as the guys who have designed this POTY but like you my VNA showed big problems...
In the end i made a test setup in a 1M dish and 10mw signal with goonhilly as my receiving reporting system.
Modifying until i got most signal across which is what we want in the end...
I saw no relation with bad swr but huge difference moving the dips up and down and wider and narrower..
In my test experience the distance between the plates need almost perfect as described and the 2 dips need 40 Mhz to 50 Mhz higher and lower as the used frequency, Than use the right cable lengt to get the optimum swr at the amplifier.
Keep in mind that 4 cm longer will influence the swr...
Even in cables over 10M but those cables are to long anyway...
Benno