If the premise is "we can learn how traffic flows from watching traffic patterns" and that's accomplished by watching where the tire marks go, then it seems pretty good to me. However, watching where the snow disappears first from the roadway isn't an accurate measure, because people drive differently in the snow.

Another example based on the premise in the article: you can find the best racing line on a race track by looking at where the track is darker, because rubber and any leaking fluids will stain the track directly under where the cars are always driving. That's where racers will be driving during normal conditions, but it'd be inaccurate to say that the race track should only be exactly that line, because drivers will occasionally go off that line to pass, avoid an accident, or participate in an accident. In both racing and normal driving, the best line isn't always the only line.

That's what I meant by the premise is good, but the execution is bad. It's just a faulty measurement.

Don't get me wrong, I think your reasoning is sound and I think I could argue your view with pretty good confidence. Your example is quite good and I like the idea of taking the totally opposite driving circumstances to prove the point.

but.....

I took the premise as "there is a lot of wasted space on the road and we can reduce it" with a methodology of "removed snow will show us what is not wasted". So, I thought the premise was hogwash, but the execution fit their twisted premise.