If used "300B area" OP, for example, 350V/-66V/75mA, it can produce "typical" 300B power, about 7W or so with Ra~3k load - it`s a probably interesting for PP. With Raa=6k we can expect, say, 14-15W in class A.
If we use more "modest" and linear OP for SE, for example 300V/75mA/-54V and use Ra = Ug*µ/Ia - rp = 54*4,2/0,075 - 700 ~ 2,5 kOhms ("standard" 2A3 value), we can expect anode voltage Ua=µ*Ug/(1+rp/Ra)= 177Vp = 125Vrms. On the anode load, Ra, it produces Pa=Ua^2/Ra = 6,25W.
If we have about 10% (OPT) losses, we can expect about 5,6W output, larger then "typical" 3,5W from standard 2A3.I tried something similar with 300B I`m experimenting. I used 760-780V ct secondary, CLCLCLC filter/AZ50 rectifier, B+ is about 440V, few V less for the driver. When I disconnected the first cap, I get L-input supply, and B+ is about 312V - exactly what we need for 2A3. We have about 12V voltage drop through the Rw of the primary (OPT), and need Uak=250V through the tube, Uk=45V.
I didn` bother to try 6B4G (I have a few, but no 2A3), I just left 300B in place - it "biased" about -51V, with 250V/56mA through the tube, Rk~900 Ohms.
E180F trioded driver gets about 310V - enough for CCS driver "work" (Ua=200V). The sound was very good, but with less "slam" - I have 89dB/W speakers :-(.
When I`ll finish this 300B amp, I`ll probably post about little changes needed to "convert" it in 2A3 amp, with "traditional" OP, 250V/45V/60mA.