These amps are worthy ! Iron is good to very good, but bandwidth gets very, very good with a bit of overall loop feedback.
Look at the amp, before stripping it, please ? The voice coil feedback arrangement can simply get changed to "more normal" values. Simply, and I mean simply, rearrange the 100 pf cap across the 3.3K cathode resistor. Keep the 3.3Ks as the 6CG7/6FQ7 cathode resistors, but twist the 100pf cap to become the feedback cap toward the speaker voice coil. Across those caps, parallel a 10K to 33K feedback resistor. So, that now unconnected end of the 100 pf/10-30K Ohms then gets connected to the white leads "directly" from the voice coil. Of course, simply lift the white lead (some versions have yellow leads for the voice coil/feedback) from those unneeded 47 Ohm/2W resistors.
One of the units I built, I used the 47 Ohm/2W at the "bottom" of each 3.3K now 6CG7 cathode resistor. That way, their junction, 47 Ohms and 3.3K Ohms, can be a great point for adding more feedback.
In fact, the feedback arrangement of the stock unit is what reduces output level. Next time I build or rebuild one, I will install a variable, 25-30K ohm pot for an easy, adjustable feedback. My user RS-199 unit, which I demo at meets all the time, sounds incredibly good with my reference Norelco(Philips) full range 8 inchers in TL boxes. That amp uses 22K feedback resistors from the voice coil, above the 3.3K cathode Rs. While some lurkers have posted the feedback is minimal, it certainly makes a big sonic improvement over stock. Plus, the sensitivity of the input is now much more usable and can even benefit from an input pot. We do not need an extra gain stage as some folks recommend.