Right there's your hard cap. Make trains too expensive relative to trucks, and, suddenly, everything goes most of the way by truck.
Unmanned trucks crossing long distances of rural America sounds like a recipe for hijacking loads.
The logistics of stopping and looting a truck involves too many parties, and ensuring that each party is following enough security protocols to not be identified via face, vehicle, or gait will ensure that only a few small sophisticated heists will ever be successful.
Logistics is complex; you'll also need to factor many things into the optimization: * both fixed and marginal costs of each mode (e.g. maintaining track, monitoring safety, wear and tear on vehicles, varying fuel costs) * constraints (due to technology, personnel, regulations, etc) * fluctuations in demand and shipping objectives * lots more
If you want to focus on only one slice of the problem... Sure, for the exact same route (meaning that a particular track has already been built), one would expect that trains are more efficient. The data shows that; e.g. https://en.wikipedia.org/wiki/Energy_efficiency_in_transport...
Not always. If you provide a value proposition that a cheaper offering does not, say speed, you can increase volume.
Highway transportation will not likely get much faster, but high speed freight via rail seems like it might have some room to grow.
