> input going directly to the volume pot and then to the first stage of the 12AX7. No other schematic I’ve looked at has the input wired like that.
That's normal for Hi-Fi amps. Sources are already pre-amplified (or hot), gain is low. Putting Volume pot in front avoids overloading.
Guitar wants more gain for easy playing. In this plan R2 reduces gain by NFB from 2nd plate to 1st plate. Your first crack is to snip one end of R2.
Now you have gain. Maybe pretty good for guitar. BUT there's hiss in the 1st tube. Working at "10" and playing steady, you won't hear the hiss under the guitar. But turning to "5" turns-down guitar but does not turn-down hiss! The thing hisses the same at all settings.
There ARE guitar amps this way. Mostly very cheap practice models. There may be a SilverTone like this.
But since guitar will not overload the 1st stage, a better plan is to re-wire the Volume pot between 1st and 2nd stages. Yes, like any Champ. It is interesting that this even uses the same 1st stage resistor values as Champ (most Fender). That should play fine.
The 1K R22 is "large", drops power. But you gotta ask: is this ever going to fill a stadium? No. Would dinking it up from 4 Watts to 5 Watts get more gigs and groupies? No. Does the 1K value significantly cut the hum (on a large speaker)? Yes. Leave it alone.
C12 is probably perished. Get a few 40uFd 400V caps.
The 6X5 is NOT a problem. Maybe if you have a fleet of bottle rectifiers working every day, the 5U4s are a bit more reliable, but I have not seen that. Leave it alone.
The grid-leak of the 6V6 is a whopping 5 Megs, and it works 10 Watt plate dissipation. Book says 1/10th that value. As the tube goes gassy the bias will drift way off. Toward melt-down! Again the 1K R22 is a mild benefit, dropping B+ as 6V6 current rises. Me, I think you have a "Tone" knob on the guitar, you should bypass the 5Meg pot with a 470K resistor. (The knob will still "tone" when turned full-down.)