I hear you on the lack of sleep CK and it gets me too. Back when sweep tubes where $5 each and tube testers were priceless, I would test them by seeing how much current they could support flat out. If they didn't turn red, a new $5 tube went in. If anyone has some of those tubes that turn red, send them to me. I'll take all of them.
All kidding aside, just because a tube can turn the plate red does not mean it's a strong tube. Many of those Chinese tubes suffer from poor quality control affecting the grid inside the tube. Alignment can be off, grid wires can break off and short to the plate or the cathode. If the grid shorts to the plate, you get fireworks from meter shunt resistors, grid resistors, grid chokes and capacitors.
If the grid shorts to the cathode (filament) you could have the exact symptoms this Heathkit is demonstrating. This effectively removes the cathode bias by shunting it to ground through the grid. That will cause excessive plate current even before drive is applied and cherry out the plates on the tubes. This will also cause the output to be very low since the drive is seeing a much lower impedance and the tubes are operating very inefficient due to the extreme bias current.