Could it be that the rods in the video are only to protect the sensors from impact?
Have the rods small holes drilled in them?
(In the 0,1 to 1mm region)
If thats the case, perhaps the whole thing gets pressurised externally with air or nitrogen with an external pressure line and produce the "real test leak" that way.
This would be the cleaning function too.
Are the wires to the potted part under the shiny dome NTC or PTC resistors?
Perhaps the heater get mounted externally if needed & the wires are only to monitor the heater.
Soldered-in fuse in that kind of equipment seems fine to me. I assume that, due to safety requirements, the PCBs are not serviceable.
I think its due to cost reduction, prevent corrosion on the terminals & prevent the fuse from falling out when the thing gets dropped.