Tested this change in a sequence of motions on a large flat plane with the same friction as the road in task 1. No significant slipping or spinning occured and driving motions are now repeatable. Did limited testing in vrc_task_1. Again, no slipping or spinning was observered when breaks were applied while still on the road. Drove off the road once and it seemed to behave reasonably.
This also seems to solve the issue of polaris drifting on the starting pen while stationary.