It would work as a compressor if V2 were normally biased near cutoff. Then positive signal peaks would cause it to conduct, discharging the 40uF cap and lowering the PA screen voltage. But negative peaks would have no effect, since the tube was practically off anyway. So basically V2 would be a half-wave rectifier with gain.
I conclude that the circuit won't work unless V2 is biased to pull only a little current at idle. It also needs to be a very strong high-gain tube to make much of an impression on that 510 ohm screen resistor. You'd need something like an EL84: the high-current section of the 7247 would hardly make a dent in it. Even so, you would probably want to increase the screen resistor too. And/or commit heresy by using a MOSFET ;-)
In short, I don't believe the circuit will work at all, as drawn.