I don't see this effects-loop; not even a place for it.
If you are looking for headroom, you know that low B+ suggests smaller dropping resistors.
14V in 4 Ohms is 50 Watts. Not unlikely for 400V and 3.4K load.
Real Bassman with 400V and 4K load is 40 Watts.
Up-sizing the tubes (5881 to 6550) won't increase output on the same load by very much.
The lower load is good for some more watts (if the tubes will pull it), and (4K/3.2K)*40W is just about 50 Watts.
V2 is an elaborated 5F6a Bassman tone driver. Which is a modded Standard Fender Stage plus a cathode follower. (The whole amp is based on 5F6a.) (Ha, it even says Bassman.) The SFS with 1.5K cathode resistor sets the plate about 69% of the supply. That's too high for a direct-coupled cathode follower, so Leo switched to 820r cathode resistor. This should bias plate nearly half the supply. 247V, 140V, that's half-enough for our purpose. And should be good for 50V peak swing into the tonestack. Taking TS loss as 10:1, that's 5V peak into the driver. The whole driver/output stage seems to work at gain of 12, or 8, possibly 5. So 5V out of tonestack "should" give 25V peak or 18Vrms at the speaker. With this B+ and load, the output stage won't do over 14V. So the stuff before it is not really restricting the output.
However 18V is not a lot of headroom over 14V.
The 247V to V2 sounds low. I thot Leo ran above 300V here.
For "MORE", a solid-state rectifier is 10%-20% more Watts output.
If that's still less than you need, double everything (100 Watt parts) and use JBL or E-V "lead guitar" speakers. Or a Full Stack (8) of hot Tens. (There's more than 2:1 difference of efficiency between "loud" speakers, and at this scale a high efficiency speaker (or array!) is better leverage than doubling the amplifier power.