Tom said;
...There was an article on the apparent problem with the 122 VFO in Electric Radio many years ago...
Actually it isn't a problem, it seems to have been designed that way.
According to the article Tom referenced, the author found that grid tank of the 122 is tuned to 1.9MC, but the plate tank is tuned to 3.8MC, as evidenced in the plate tank pic below.
This was done because the Viking II for which the VFO was apparently designed, and subsequent EFJ transmitter oscillator circuits, had band switch selected 1.9MC plate tanks. This meant that there would be sufficient 1.9MC fundamental content in the VFO's output, that the crystal oscillator stage could clean it up and deliver the required 1.9MMC signal to the following "multiplier" stage in the transmitters.
---------------
If you want to clean up the 1.9MC you would need a total of about 500pF for the plate tank to resonate properly, but then, there would be virtually no 3.8MC content.