I would think that if you could record a test record track, like the 1kHz or 3kHz ones, that might help.
Usually with an ADC (Analogue to Digital Converter.) or with the ADC built into a sound card or microphone input on a PC.
I suppose if it good with the previous/other arm, then I assume we are thinking that the platter speed is not the cause.
From memory the long arm is something 19gm effective mass, and the 9” was 13 or 14 grams. So I don’t think adding the Schroder brass spacer to increase the effective mass is going to be helping.
Does the arm have any visual oscillation ?
If that is not it, then I would have thought that any resonance would have presented as something different than a warble.
Have you reached out to Frank via email? He is pretty responsive.
And I would be interesting in knowing what is happening.