This tube's cathode draws 90 Amps at 5 Volts.
The wire on the cathode choke looks a little wimpy for that kind of current.
Have you measured the cathode (filament) voltage *AT* the tube socket? Excess voltage drop to the cathode not only holds back power, but you risk damage to the cathode. Running RF with the cathode at a temperature that's to low screws up the vital layer of thorium. Without that, the cathode becomes "soft", and saturates at a lower current than it's rated to pull.
Best example of this is with the Black Cat JB2000 amplifier. Anyone who runs it from 120 Volts will see the filaments dim when you key the thing. Wears out a set of tubes in months of heavy use. The power just drops off, bit by bit until you can't even get half what it used to deliver.
That model is vulnerable because it uses only a single transformer. Makes the filament voltage drop more than it does in a design with a separate filament transformer.
Running it from 240 Volts removes this risk. And simply turning the power down to half would probably prevent the "120-Volt problem".
Never have seen anyone try that solution.
73