You must select the volume and tone resting frequencies so that harmonics of the tone oscillators (multiples) will not interact. Example: A multiple of the tone oscillator is not close to the volume oscillator frequency. If you don't do this, you will get all kinds of garbage in your sound. The basic design works well BUT it must be tuned correctly. All the tubes on the sound chassis must be shielded types. Use octal tubes! Substituting different style/number tubes will alter the design in terms of transconductance, impedances, MU, etc. Stick as close to the original design as possible or you will run into issues. Mark Keppinger is still around on Facebook so you may be able to interact with the Master.
Did you tightly twist your filament windings? You may also wish to investigate a 'Hum bias" circuit if you have AC hum riding on the output waveform. Remember, these tubes are AC heated filaments.
Orient the chassis mounted coils 90 degrees to each other to prevent inductive coupling or they will interact. An oscilloscope combined with a freek counter are essentials. IIRC my tone freek is around 196 KHz and the volume freek is 443 KHz.
Also, look up "concert tuning" and "bass tuning". Some of caps in the resonant circuits are changed out. I find that concert tuning gives me the best octave range.
I use full size RCA reproduction aerials so the hand response is excellent. If you use small whip aerials, there is less surface area for interaction. Capacitive coupling is all about surface area and distance and of course the dielectric constant the insulating medium (air in this case).
Fleep