Yes, this guy is running around in circles, when the problem is staring at his nose.one thing he did not do was change the bias voltage or lower the B+.
He could have looked at the datasheets and found that a good compromise between EL34 and 6L6 at 450V B+ is between 5k and 5.6k.he might have got 5K:8 or 5.2K:8 instead of 5.8K:8.
I believe Mesa expected it to be driven with one of their tube pres like the formula, still no reason for two resisters, low to cut the signal?Less than 10k ohms is low for a guitar amp , then again they probably expect the power amp to be driven with solid state gear ahead of it . You could change it up to 1M if you wanted to connect a normal anode load 12ax7 stage ahead of it , you will get a little extra hum from the AC heater supply into the grid also.
The 220k ohms in series with the grid of the power tubes looks a bit odd too .
There are a few old Fender cathode biased amps connect the heater centre tap to the power amp cathodes and across the cathode resistor to ground , not an option on a fixed bias amp like the 20/20 .