Vinyl Asylum: I agree with Baerwald being mathematically attractive, but... by travisty

Distortion is distortion. What difference does it matter where on the record it occurs.

I am not ready to stipulate this. For two reasons.
1) Tracking Angle Distortion (the one we are talking about) does not care about where on the record the music is, and it assumes a constant velocity of signal. In the real world, some music has its most dramatic and complicated parts at the end of the side. I'd rather have less distortion there than in the part where it is just a single violin playing. This is not necessarily the case, but it may be dependent on the music one listens to.
2) all the other distortions out there (discussed below) which are almost never discussed, which have nothing to do with the arm, but which ARE groove-radius dependent and frequency-dependent, and which may be amplified by tracking angle error.

I agree there are mathematical attractions to Baerwald/LofgrenA. On a practical basis however, if one has standard mounting distance according to a standard IEC template, the EFFECTIVE best alignment for any given record made in the past 25yrs could easily be Lofgren B because so many of them have their last modulated groove radius even before the Baerwald inner null point.

However, all things said, if you get any one of the alignments spot-on, you are probably doing OK for most modern records. Lateral tracking angle distortion is mostly second harmonic, which is not too bothersome, and I have yet to figure out how ANYONE can accurately measure vertical tracking angle distortion given all the inputs which go into it, and the fact that music is definitely not a sine wave.

Stevenson's alignment produces greater distortion in-between the null-points for a longer period of time.

Agreed.

Furthermore, by placing the inner null-point exactly at the innermost groove, RMS distortion over the last centimeter of modulated grooves is significantly greater than from L�fgren's "A" alignment.

I am certain I have not read deeply enough into this subject, but as far as I can tell (and not knowing the source of the constant in the pivot-to-spindle distance calculation for Stevenson, I could easily be wrong), Stevenson null points are not generated off groove radii like the Lofgren A & B are. That means your conclusion is not necessarily the case as far as I can tell - it will depend on where the last centimeter actually is.

In my calculation, Stevenson does have higher average tracking distortion for records where the last centimeter is the 60-69mm from the spindle (i.e. an IEC record and alignment), but has lower distortion than Baerwald/LofgrenA for records ending at 57mm, even if Baerwald/LofgrenA is set to DIN groove parameters.

OTHER DISTORTION
As to your first point that "distortion is distortion", that is one I am not ready to stipulate, not only because I am not convinced music across a record is necessarily equally distortion-sensitive.

Tracking angle distortion (the one we discuss when we debate Baerwald vs Stevenson) is actually quite small compared to other distortions which are particularly available on stereo LPs - such as vertical and horizontal tracing distortion or pinch effect, distortion due to VMA/VTA differences, or assumption of lacquer resiliency or vinyl resiliency (though vinyl resiliency. Furthermore, there is high-frequency loss at inner grooves (though this is heavily dependent on stylus tip radius), which is modulated by both frequency and signal amplitude (higher frequency, lower groove radius, lower signal amplitude all serve as a "weight" on reproduced HF signal according to the math I have seen. Heck - there's even the distortion which comes from the hole being off-center. All of these distortions exist above and beyond tracking angle distortion, and they increase as groove radius decreases, and as frequency increases (the high frequency level loss is not so much distortion vs the appropriate curve, but that the curve itself is lossy, substantially less so with very, very small stylus tip radii). Furthermore, distortion of distortion can be, but is not necessarily, non-additive, suggesting that while tracking angle distortion itself may not be a biggie, tracking error can be because it aggravates some of the other distortions.

Some of those are second-order harmonic distortions and some cutters used Neumann tracing simulators cut off-phase from the program in order to cancel out the distortions when presented in playback, but I have no idea as to how well that was implemented and/or how often it has been used, so I still have a fair bit of learning to do in order to feel comfortable that indeed, "distortion is distortion."