Determining what is causing the shift in frequency response has been most interesting. One of the references in the Kokusai patent is another patent assigned to Clevite Corporation dating from 1959.
https://patentimages.storage.googleapis.com/99/b4/ca/7bf470f03be104/US2877432.pdfThis patent provides details about the ceramic transducer including that it is cut to dimensions that resonate at 455 kHz. To examine the transducers more closely, the filter was stripped of all the old wiring. In the process one of soldered encaps came off a transducer.

It was soldered back then the resistance of the transducers measured as a function of frequency. This image is of the response of the transducer that had the endcap come off.

It has a singular response but very low in frequency.
The other transducers response is below,

There are multiple response peaks with 455 kHz between them.
Working with the first transducer with the clean response, by carefully positioning the endcap the frequency of the response has been moved up to 414 khz but still too low. This seems like a very delicate operation to get the frequency correct. It may require a fixture to do it with sufficient accuracy.
Replacement wire for the filter came from a damaged stereo earbud cable.