[Chas]

Hmm,

Seems I am guilty of allowing overthinking of this project  

I have posted this snip of the output area, I want to post more from the power supply for reference but I don't want solutions to get carried away...

   

Look to the top/right corner, see the cone icon, immediately below represents the voice coil, below that is the hum bucking coil... It is that common junction that must be restored to "help" reduce the likelihood of having to swap "voice connections" to reduce hum. In other words this is a proper connection leave it alone. The hum buck coil is mounted on the same pole piece next to the field coil. Therefore, for this speaker, the "voice-coil" impedance is the sum of both coils for determining OPT output impedance

IMHO an output transformer for 7 to 10K plate to 4 to 8 ohms impedance at roughly 5 watts should be well within the "ballpark" for this radio.

Any "precise engineering" will give precise headaches, this is a technicians repair job no more...

The plate/B+ wires on  new OPT will be identified, connect as shown in docs for the OPT. The "voice-coil" wires will be either all black or coded, none the less make temporary connections to the series combined voice-coil/hum-buck ends, then operate the radio, if there is residual low level hum with volume control turned down, then swap those two transformer secondary connections. If the hum decreases keep and make permanent, if not the original connections are the best it can be...

This radio uses voltage drop from a tapped or separate string of resistor from the center-tap of the power transformer to chassis/ground connection. This group of resistors must be intact, not shorted or seriously changed in resistance.

Look and see that next to the power transformer center tap is a wire connection, this is the most negative voltage source and via a 1 meg resistor provide bias to the 6F6. If this bias is too low to zero the tube will conduct heavily, meaning an excess of plate current, it could have been a cause of OPT failure...

Another loss of negative grid voltage for the 6F6 is the signal coupling capacitor from the plate of the previous stage to the grid of the 6F6, if this began to leak B+ it would allow the 6F6 grid to become more "positive" and again conduct heavily.

So, the OPT gave up the ghost, despite theoretical analytics of mine there are other external causes that a technician must look for a successful and robust repair.

You have a copy of the complete schematic so you know the areas of the power supply I am referring to.

As for the combined tone capacitors, use at least the voltage rating specified but know that increasing the voltage rating of this capacitor in a common metalized unit, will not prevent its ruin from spikes, as it is the metalizing that fails and not from a puncture of the dielectric.
FWIW use a common metalized in that tone position to get the radio up and running and plan to replace the cap with a dv/dt rate cap in the future, avoid listening during a thunderstorm until the cap is replaced. Though a new OPT will be much more robust, better insulated, why chance?
-------------------------------------------------------
It is always good to retain the original field coil speaker along with the close approximation of the filter cap values.
-------------------------------------------------------
Do, have a replacement 6F6 handy as MrFixr55 has stated they are easily roasted, I have been guilty of such cooking too.
-------------------------------------------------------
Ron's suggestion is also an ideal solution for an OPT because the manufacturing of such audio distribution transformers is still very competitive, the prices are quite low when compared to a conventional OPT of limited manufacture.

GL

Chas