Trains accelerate slowly. Adding stops non-linearly increases travel time [1]. A few extra stops might thus make total travel time by train meaningfully more than by air.
[1] https://blog.bimajority.org/2018/10/20/interlude-physics-of-...
> There may be reasons to add or update infrastucture ...
As long as we're talking in abstract we can construct scenarios where one is faster or better. However my read the original comment was that that adding a new connection involves adding new infrastructure, which isn't always true.
The single trip I took by rail in the US we spent most of our time just waiting on other freight trains to pass us.
On short distances a railway is simply less efficient than trucks.
Also companies have optimised their storage and logistic chain to expect for example one truck each day instead of a train each week and thus are using the road as storage facility.
Part of it is that the Northeast Corridor has to pay for basically the entire rest of the country where Amtrak almost universally loses money.
The other is that Acela specifically is priced to be more or less just competitive with air on the BOS-NYC and NYC-DC legs. (Acela doesn't really work for the most part to travel the entire length of the corridor.)
I almost never take Acela if I'm paying myself because the electrification work done for Acela made the regular regional train only about an hour slower for the BOS-NYC leg. (They used to have to switch engines in New Haven.)
Passenger rail is pretty much the direct opposite, requiring fast acceleration and speed. So the two do not mix well (freight trains will slow down passenger trains.)
The only country that operates both high speed rail and freight successfully, by China, literally built a whole separate new high speed network to reserve for passenger trains and to leave freight the freed up space on legacy lines.
A bit like what the US does for roads.
https://specialty-freight.com/2021/06/us-requires-railroads-...
"This April, the US Senate introduced the Rail Passenger Fairness Act, legislation that requires infrastructure managers to prioritize passenger trains over freight trains. Infrastructure managers will be brought to court for not complying.
[...] the US is giving Amtrak, the nation’s main passenger carrier, the ability to take infrastructure managers to court for not providing preference to passenger rail operating on their rail lines."
Apparently the incentives are not there for that, and should be reexamined accordingly.
The thing with air travel vs. trains is that with air travel you only need infrastructure at the origin and destination. This makes it fairly easy to add new routes. By contrast, for a train you need infrastructure all the way from origin to destination, necessarily crossing other types of infrastructure (e.g. roads and canals) which requires the construction of bridges and/or tunnels. The land also needs to be acquired and, depending on length, will likely require dealing with hundreds or thousands of existing land owners.
If there is an existing rail line, that may help, depending on spare capacity. Adding additional tracks may also be slightly less difficult than constructing an entirely new track, assuming the existing property is wide enough to accommodate the additional track(s).
A port for an airship is probably even easier to realise than a normal airport, as you don't need a long runway.