I’ve very successfully used a circuit board with smd parts for an EMC barrier. The goal is to have virtually no lead length between the shielded and non shielded volume.
This concept adapted to your device would look something like this:
A small board that is mostly ground plane (GP).
This board will be bolted, soldered etc to the housing with the below mentioned via’s in an opening in the metal case.
The filters would be on the “inside” side of the board, surrounded by GP with the output of the filter terminated by a large via.
On the other side (still mostly GP) the via would either accept a wire, or a connect to an SMD header.
This design passed Military EMC requirement in a product that measure a precision RTD.
Alternately, if you didn’t want a separate board you would probably get most of the EMC benefit if you used the same concept on a horizontal board with the Filters on one end, the outputs terminating in a 90° header configuration.
Thanks for the comments. I am using a Murata FT part and will see how it goes. There is not only the conducted emission issue, but radiated which I am trying to address as well. My latest thought is to have the enclosure end panel be a PCB with exposed Gnd where I might solder the Gnd top pads of the FT cap. I also plan to fill the enclosure with RF absorbing foam in an insulated covering. This is like meeting DO-160 on a budget (8-).