I looked at your MicroSquirt Tooth Logger logs and think I understand most of what I'm seeing, but I'm still puzzled by your original picture where it was running OK and then lost sync at ~2400rpm.
Below are some example CPS VR oscilloscope voltage measurements from my wagon (LH2.4). Notice that there's a fair amount of peak-peak voltage variation within a single rev, and that the peak-peak voltage goes up as the rpm's increase. The once around 60-2 "sixty minus two" missing tooth region is clearly visible (click pics for bigger versions):
1) cranking ~200rpm, ~0.9V peak-peak
2) idle, ~850rpm, ~3.3V peak-peak
3) rev'd up in park, ~3100rpm, ~10V peak-peak (I'd expect ~20V peak-to-peak at ~6200rpm)
The problem MicroSquirt has with the 60-2 VR sensors is the overall range of peak-to-peak voltages from cranking to max RPM (~0.9V to ~20V with LH2.4, maybe more with MS). Without the 10K resistor to reduce the peak-to-peak voltage, the MS becomes very sensitive to flywheel finish and height variation in the missing tooth region - at max RPM, a little variation can generate enough voltage to look like an extra tooth. The flip side is that the 10K resistor also reduces the already small voltage seen during cranking, and may cause starting difficulties.
Below are MegaLogViewer screenshots from your 2022-05-22_11.08.12.csv Tooth Logger file. The graphs are the time per tooth in milliseconds. To convert to RPM, simply calculate 1/tooth_time, for example, a tooth time of 0.5ms is 0.0005 seconds, which gives an rpm of 1/0.0005 = 2000 rpm [the full equation is RPM = 60seconds/minute * rev/60teeth * 1/0.0005seconds_per_tooth, but the 60's drop out]
1) running properly at ~1850rpm (0.54ms) - notice that there are 57 normal pulses then a 3x longer pulse (due to the missing 2 teeth getting lumped into the next tooth's time), and that there are no unexpected skipped or extra teeth.
2) idling at ~900rpm (1.1ms) - notice that there is a little twice-per-rev speed variation due to slight slowing and accelerating before/after TDC and BDC. No skipped or extra teeth.
3) cranking at ~200rpm (5ms) and startup to ~720rpm (1.4ms). Notice the pronounced speed variation during cranking, and the somewhat inconsistent tooth times during the slowest regions, but no skipped teeth. After startup, notice that teeth are being skipped, resulting the long tooth times at non 60-2 spacings - this is bad.
4) cranking but no start - notice that there are lots of skipped teeth, resulting in long spikes at non 60-2 spacings. My guess would be that the VR signal voltage is too small during cranking for the MicroSquirt to reliably detect. Changing to a Hall sensor should fix this.
Let me know if there are questions or confusion (I didn't proofread this).
One more question - what version MicroSquirt do you have? The newest AMP EFI version has internal dip-switch selectable pullups for Hall sensors, which would cause problems if enabled with a VR sensor.
-Bob
edit: here's a link to the MicroSquirt V3 VR Sensor documentation
link