Well, it isn't conclusive with parallax and all, but without precise values on exactly when those verticals occur, it looks like the signal itself might be wonkey (lot of slop in the speed signal on these cars?!?)
if I simply use paintbrush zoomed in and visually compare the widths, there does seem to be a noticable variance. Maybe I need a certain number of pulses or something? Maybe the vss cam has multiple lobes and they aren't terribly precise? Other thoughts?
I suppose I could lift a wheel and hook up a buzzer to the vss line and see...