These problems only occur when the load get close to the effective maximum load. It's easy to look for blame in driver behavior, but I think misleading -- it's

These problems only occur when the load get close to the effective maximum load. It's easy to look for blame in driver behavior, but I think misleading -- it's like taking that first crystal to form in 0 degree water, and blaming it for ice formation. You may be able to reduce crystallization (and thus sneak in a couple more Joules) by distilling the water (adaptive cruise control) or cleaning your glassware (on-ramp metering or "do not change lanes" signs), but really you're just fighting the statistics of phase changes. Note that I don't have a ton of supporting evidence for this sort of thing. One imaginary project that I would love to see if I had the time or the patience for deep research is to just gather a data set and make it public of anonymized traffic behavior -- over a large stretch of highway, have periodic cameras that track every car by license plate (if possible) or car shape, aggregate and remove the identifying data, and output a data set that allows you to slice it by time or by moving segment to directly observe these things.