>
class of operation depends on the bias conditionAND the grid-drive you apply.
Any amp can be driven into grid current.
If you do not plan for grid current, it will sound awful.
If you do plan for grid current, it may make more power.
Datasheet PDF pages 12 and 14 (original pages 8 and 9).
Take the right-most column, 300V on plates.
Without:
Vg1 (bias) = -26V
Vg1g1'p = 52V (26V peak signal each grid)
With:
Vg1 (bias) = -25V
Vg1g1'p = 75V (37V peak signal each grid)
In the first case, you use it so the peak grid voltage goes barely up to zero but not positive. Grid current is nearly zero over the whole cycle. 22 Watts out.
In the second case, you force peak grid voltage up to zero and then higher, to +13V. Grid current is nearly zero up to zero-grid, then rises near 3mA average but 20mA peak. 37 Watts out.
When grid is zero to negative, the grid impedance is high, essentially the 50K max grid resistor. When the grid swings positive it quickly drops to 1K and then toward 600 Ohms.
You can't drive such a load with a small triode and R-C coupling. The small triode won't make 20mA into a load, and the R-C into a load that only sucks part of one way will quickly charge-up and bias-down the stage. Grid current demands a driver which is a small power tube, and usually a transformer also.
And when you do your budget, you might decide that a larger power-tube (or the same with a different condition) working no-grid-current can make the required power cheaper than a with-grid-current scheme plus the beefy driver.
With-grid-current makes sense mainly when there is a not a bigger tube (or higher condition) available at good price. And especially when you are going to use a driver transformer anyway. In 1938 it was very common to drive 6F6 and 6L6 class AB2 to get a little more power without going to more parallel tubes.
Grid current makes much more sense in RF work, because tuned circuits are essential anyway and can usually be tapped to provide the high peak grid current from stored energy.
For a guitar amp you almost certainly want the NO-grid-current conditions. Only the 300 Watt SVT and Fender 300 ran to grid current; 99.9% of our favorites run R-C coupled small triode drivers.
Unless QQE06/40 are super cheap, I don't think they are an ideal guitar amp tube. Yes you can get 50 Watts with no grid current, but at super-high 600V supply and 12K load. You also need a 250V screen supply with good regulation from 1mA to 23mA. The heater requirements are the same as two 6L6, which in modern forms can make the same 50 Watts with much lower supply and load values, and are probably easier to find in a strange town. The extra heater power of EL34 is not large and they are sure readily available and ample for 50 Watts. Also the Septar socket and the two exposed plate-pins.