That is a lot of dissipation for the 300B.
The filament supply isn't shown, which is odd.
The EL84 as a pentode is a great driver.
If you have an old tube regulated bench power supply, you could get a decent approximation for the sound of the amp by lashing together one channel. There is a significant omission in the power supply schematic where a 6AS7 and the EF86 share a heater winding, but no specific biasing up of that winding is shown.
I'm having a tough time swallowing the 6.3V/10A you're going to need for the regulator section.