Let's put some physics and math to it for Ralph.
1. Skating force is always in the same direction, never changing from + to -. If we denote the inward radial force (skating force) as positive then it remains positive.
2. The magnitude of the vector changes in relation to the error in tangency of the stylus to the groove. Since the angle relative to tangency of a pivoted arm changes then the magnitude of the skating force changes.
3. Though skating force always remains positive (in my convention of positive meaning toward center) the DEVIATION in skating force goes from positive to negative to positive across the record as it is played from the outer groove.
However, the deviation in his force is pretty small, given that skating force can vary from record to record a bit anyway, and given that the antiskating mechanisms of arms are not perfectly linear.
4. The magnitude of change in skating force for an arm with a 23 degree headshell angle would be +7%, -4%, if the arm uses Baerwald geometry. The percentages are deviations from when the arm is at true tangency (null points).
5. Let's look at the magnitude of those percentage changes. Based on my detailed and repeatable experiments to determine actual skating forces I am confident we are looking at deviations of approximately two HUNDREDTHS of a gram! A small Post-It note weighs TWENTY times that!
Now, this is where experience can be helpful. Try as I might, I cannot reliably detect a sonic difference by changing the antiskate force by those amounts. I cannot detect a tracking ability change by changing antiskate by those amounts.
4. With poor tracking pivoted arms, yes, tracking is worse at the center of the record. But, since the best arm I can find is free of this, I have to conclude that there are many problems at work with those arms, including geometry (one low cost, ubiquitous brand uses their own geometry that nobody can figure out), resonances, and other matters that I'm not about to spend time figuring out.
Based on all of the above, I have concluded that variations in skating force across the record due to geometrical considerations are not a material concern in the sound reproduction of a pivoted tonearm.