for a race car you could still use my rpm method, just calculate it on the straights, algorithms get a little tougher, but only slightly. Hard to accommodate for tire spin.
you may be able to electrically isolate a tread block and check it’s resistance somehow, to get an exact wear percent… not positive.