Well, based on the answers I got from you guys, I guess this theory is false.That's wasn't my answer or Nsgarch's. I said the theory was true but that attempting to apply it would be ridiculous. Nsgarch elaborated by explaining why matching VTA between individual cutting lathes and cartridges would require too many cartridges and would be virtually inaudible anyway.
Gregadd,
Put a stick of soft butter flat on a plate.
Hold a sharp, pointed knife straight vertical, with the edges of the blade oriented across the stick.
Lower the point into the butter and drag it the length of the stick, moving it left and right as you go.
Now rotate the butter to put a fresh side on top.
Hold the knife edges across the stick as before, but angle the point sharply either point-forward or point-back.
Cut a modulated groove as before.
Compare the two grooves. They are clearly different. The modulations cut with the vertical knife have straight vertical sides. The modulations cut with the angled knife have angled sides. The only way to accurately re-trace either groove is to use a blade with the same size and shape and hold it at the same angle (SRA). Any other angle will cause the blade to slur past parts of each modulation, creating playback timing errors.
We adjust arm height for each LP. The changes we hear are consistent with SRA-matching theory. Some people do not hear the differences as readily as others. That's normal, but it doesn't change the shape of the butter!