My lazy googling suggests that 40% of freight movement is via rail in US. I can't find a full breakout, but the same site goes on to claim rail is only 1.7% of GHG, which means most of the rest is traveling by road, since waterborne freight is generally more efficient than rail.
In my head, semi-trucks (and maybe all cars?) only need 50-100 miles or so of range for low-speed (where batteries shine) warehouse/ports/downtown deliveries (or garage-charged commutes?).
Almost every trip I've ever taken beyond ~20 miles included at least some portion on a highway. Electric interstates could allow NY to LA in one shot and pulling off the highway with 100 miles to make the round-trip to wherever a load needed to be dropped off.
If attaching and detaching was automatic, there would be no need to electrify onramps or intersections or underpasses or anything expensive, just find the cheapest, straightest, easiest to electrify 80% of roads and slap up wires there.
If paved roads, then you're forgetting the enormous wear that heavy vehicles put on road surfaces (which would get much worse as labor costs were taken away as a downward pressure on road freight volume.)
If rail, then I think you could see some version of that as a logical upgrade to a high-utilization rail network in a high-trust society with a strong safety culture. I see the US becoming lower-, not higher-, trust.
Towns and cities are too large and too dense and too insanely expensive in terms of property prices to get new lines into or even close to.
(This might be less of a problem for freight compared to moving people, but many lines are mixed freight/passenger use, aren't they?)
In reality, such a series of actions could only exist in a society that got seriously focused on moving away from private auto, and that means density would increasingly be in style. As soon as the NIMBY floodgates came down, property prices would plummet in all but the smallest cores of desirable areas, since it would be readily apparent that you can't build all expensive areas to the level of intensity that justifies insane land values seen in sprawling cities.
Again, we could, but I recognize that we don't. Let's just not say that we can't.
This reminds me of many high GDP parts of Europe where there are towns of 500-1000 people every few km, connected by rail and busses.
But, yes, you would expect to see significant clustering. Nonetheless, the most effective clustering since the advent of indoor plumbing is the cluster of clusters, the city, and I doubt that would change. The connectivity is just so much better than a lineal expansion of small towns along a corridor.
I agree time wouldnt change that much, but presumably that time would translate to greater distances with better throughput and speed.