[Radio Wrangler]

The other way of getting RF isolation of coupling via the heater system is to have decoupling in the heater system. Most usually capacitors to chassis, less often including series chokes.

This increases cost. and the shunt impedances aren't zero, but then the single wire versus chassis approach isn't perfect either. It might get a shunt capacitor or two if trouble is hit.

It gets problematical with series string heaters in AC/DC sets. If the chassis metal is RF ground as well as neutral, shunt capacitors can be used, but they have mains AC across them and will take current, distorting the distribution of heating amongst the valves if the capacitors are large. If it's one of those sets with isolated chassis and hot circuitry, you can wind up with more safety-critical capacitors bridging taps on the heater string to what should be mains-isolated metalwork.

I avoid AC/DC sets.
I prefer single wire for radios and balanced heater supplies for hifi.

In sme cases like calcode RF amplifiers in VHF/UHF front ends, you'll see the top valve with twin RF chokes in the heater feed, possibly done as a common mode choke. In this case the top cathode must not have capacitance to ground, and the heater/cathode capacitance is too much.

Another place where heater circuitry had to get creative is for thermionic diodes in AC voltmeters, particularly RF ones. Sometimes the operational circuitry had to be designed around the limitations of the heater scheme.

David