Rail has the potential to stay even with aircraft on shorter routes: Boston-DC (& in between), Dallas-Houston, L.A.-SF or L.A.-LV, Detroit-Chicago-Cleveland.
Can it be done after WW2? IMO, yes: take off the taxes on the land under the railbeds, so railways aren't subsidizing their competition. (I'd add, give even $0.50 to rail for every $1 into Interstate.) That, IMO, goes a long way to keeping rail competitive.
The biggest issue is the grade crossings, & those need to be addressed; getting USG money to replace them all would be a big help.
Getting rid of some of the plain stupid regulations would help, too. Everything from requirements to sound horns passing through towns to track curve radii different from Europe (are European rail curves really unsafe?) to requiring passenger cars to (notionally) survive the impact of a loaded freight train.
(IDK what can, short of a nuclear containment vessel, & AFAIK, nobody's proposing putting them on railcars with passengers in them.
) That would help with performance.
Dealing with the land requirements for track straightening, without serious eminent domain issues, is one IDK enough to comment on. Straightening would be a good idea in many places.
It is possible, IMO, to raise speeds to a 200-250mph peak speed without totally relaying all the short/medium-line track, or, at least, not all at once.
Getting service between Chicago & Denver & San Francisco/L.A./Seattle competitive with even a DC-6 or Connie is a non-starter short of replacing with maglev, AFAICT.
All those Walmart trucks you're seeing? They're going from a distribution center to the store.
All the semis I see on the highway aren't, & that freight can still be carried by rail, & more cheaply than by trucks. It would put a lot of long-haul truckers out of work (& my uncle used to be one of them), but I'll bet goods would be cheaper in general. Energy consuption across the U.S. would be down, too.