To make matters complicated in practice, gas in the US (since that's where this argument makes sense) has long been considered the holy grail of the economy, and gas taxes have not been raised to account for inflation since 1990. Since the 1980s, the US has politically favored cars and airplanes over trains and buses. Roads are never expected to make a profit, though most transit in the US is. Aviation has had things like the Essential Air Service that offer subsidies for rural air routes and carriers that serve rural airports, whereas the closest thing that transit has to this is Amtrak mandates to service cities; these are hamfisted mandates that Congressmen stuff into spending bills to make their constituents happy but don't actually create the kind of healthy market you would need to offer quality service. Moreover Amtrak track is often leased from freight carriers _because_ America apportions less money to rail than it does to roads and transit, so Amtrak trains often must yield and wait for freight trains to go first. That and the fact that externalities from emissions aren't accounted for mean any alternatives to the plane and the car in the US are at a heavy disadvantage.
That's one angle to look at it. As I said, in a country that has more or less static travel routes, railroads will win over airports for the medium range. Because sooner or later the investment into the many tracks will pay off. But long-range and dynamic routes favors the airport.
But my point was that taxing energy usage heavily will bite back. You cannot build infrastructure if energy is too expensive. So either you have cheap renewable energy, for cars as well as trains, or any mechanized transportation (and much more) will become ridiculously expensive.
> gas taxes have not been raised to account for inflation
Are gas taxes absolute values in the US?
A 10 km commute by train typically takes 30-40 minutes door-to-door. For driving, the normal time might be 20-30 minutes, but the variance is often much higher than by train. If driving is consistently faster or slower than that, people tend to switch between train and car or otherwise change their commuting habits until congestion is back to "normal".
For comparison: It takes 25-30 min to ride by bike, depending on effort, weather, etc.
Unless you propose to decide top-down which projects deserve cheap energy and which don't, you cannot avoid the fact that building roads and allowing car ownership is under certain circumstances more economical than building lots of railroads.
On top of that, imagine how commuting 10km by train vs. car work out: By car it's a 10min ride, assuming no congestion. Whereas the door-to-door train trip will take the better part of an hour.
There is a similar advantage for flying. Yes, kerosene is comparatively cheap. But flying has an even more important advantage over trains and that is flexibility. The network of airports multiplies the number of potential connections whereas train stations can only lead you along railroads. In a region with fixed, medium-range traveling routes, say France or Japan, the train wins, otherwise the plane is just more efficient.