the two big filter caps look like they are in parallel with the old "wets"

That does look suss. I'd re-stuff the original -- only because I quite like doing that.

After looking closely at the under-photo, I'm going to suggest the valve lineup I gave in post #21, except for the output valve.

It has its own heater winding on the transformer, so the voltage needs to be measured. If it's 2.5 volts, use a 2A5 valve, but if it's 6.3 volts, use a type 42 instead. The 2 valves are otherwise identical.

The unwound resistor is the output valve's cathode resistor, so you could initially replace it with a 220 ohm, and at the appropriate time adjust it to get 16.5 volts on the cathode.

The choke is directly connected to the transformer, so perhaps it is in the earth (centre-tap) line instead of the usual place. It's still somewhat suspicious though, as a radio of that age would normally use the speaker's field winding instead of a choke. I think Marcc is correct in thinking someone has been fiddling.

It might be an idea to use daylight, the desk-light or the tissues (perhaps use aperture priority, or the setting for daisy's) and see if you can get a photo of the transformer & choke. If you can get side light that will do better at showing the numbers. The odd terminal makes me suspect a shielded transformer & there's another clue. Shielded types on a proper circuit always show an earth onto the core. What voltages are on it can be useful.

I have seen Midwest and an AWA with both an earth choke and in one case an electrodynamic woofer & tweeter plus two rectifier valves, one in each phase, speakers driven by 4 6F6's.

It could be that didums was trying to get rid of the hum caused by being too uneducated to appreciate the filter error made? If its not original?

Overloaded transformers do cause hum & sometimes the hum is smell from the thing cooking.

I think I can see enough to draw you a circuit if this post is still running.

The output valve is probably a directly-heated one, hence the separate winding on the power tranny.. I can't recall the type number but it's older than a 2A5. I may even have one.

Great project!

OK, my original guess for the output valve was a 47 on the basis that there was a separate filament winding on the mains transformer and the 47 is directly heated..
However, the socket in this chassis is 6 pin, the 47 has 5 pins.
My 2nd guess then is the socket was changed (possibly early in the set's life) to use a 2A5, the heater voltage is the same as the 47.

I'd LOVE to do a resto on that chassis!

A circuit should turn up, if it can be identified. One missing valve would be a rectifier, Valves are not too hard to come by usually, most bits like cas and resistors are still easy to get, nice work in progress. Getting it up and running shouldn't be too hard for the right brain, with the right exerience!

I can draw you a circuit using Protel (I have added some valves to the library!) based on what I can see and autodyne circuits of the period. Just say the word, it would take me an hour or so because of the need to cross-check everything.

I suspect the HT winding CT choke is original. Added to reduce hum level I suspect.

Yes I would re-stuff the can electros. Wish I could have got my hands on a couple of those when I was restoring an Airzone Radiostar about 9 years ago. Originals were missing. I used Berocca cans, paint stripped, to make replicas. I created the base from a large round-head bolt with a hole drilled down the middle. Result was very convincing!