Counting all the factors, I've seen reports of several studies that come up with multiples in favor of train efficiency in the 5-10 range, so the original calculation here isn't out of line even if kind of back of the envelope.

Another thing to consider is exhaust emissions - locomotives tend to be dirtier than trucks per hp (emission controls for trucks have been around longer so the overall fleet is cleaner), but with much less hp and fuel use/ton of load (yes, even counting the higher tare of rr cars) it's quite possible for the train to be cleaner on suitable runs. Not all trains are cleaner than trucks would be for a given run, but in many useful cases they can be. And for Greenhouse Gas emissions the trains almost always win for runs over a couple hundred miles - far more fuel efficient in line haul as the originally calculated multiple shows.