Getting people around quicker is part of that (many models use a cost-per-hour for citizen time on the train). More efficient connections are part of travel time, and a simulation of a large number of potential journeys on some proposed infrastructure would show that.
That means that when government is considering how to allocate $X of budget, they should naturally end up choosing projects that help align the schedules, even if nobody explicitly tried to do such an alignment, simply by choosing the projects that are best value for money.
They still ended up approving many negative projects, even while there were other cheaper/better located transit projects with more benefits. The political cost of cancelling those projects was higher (ex: Mayor has campaigned on getting it approved and there was no way he would get back on that)
It's still interesting to see it and calculate it, as a matter of transparency, but its still not a guarantee of success.
https://toronto.ctvnews.ca/costs-for-scarborough-subway-and-...
For example, this project: https://www.octa.net/LOSSAN-Rail-Corridor-Agency/Central-Coa... doesn't really change average transit times, but would allow two or three trains to "start" each day at the farthest north.
Of course US passenger rail is in such a state that there's almost always very obvious low-hanging fruit.
[95]: https://www.octa.net/LOSSAN-Rail-Corridor-Agency/Overview/
They may do this, but whether the number they derive has any kind of meaning or value is another thing altogether.
This is basically impossible to do short of shutting the system down for year to measure the loss (unthinkable).
And even then, you'd have a measurement that doesn't mean anything because it couldn't measure the opportunity costs of the path not taken when the chocie was made initially.
As a matter of fact, I'll go one step further: I don't even think the question makes sense.
Economies are chaotic systems, and even if it is sometimes rather clear after the fact that a specific choice leads to overall positive outcomes (such as choosing to implement a high-quality public transportation system in Switzerland), actually measuring the "gain" is basically impossible. Worse: it's an ill-defined question.
whether the model's assumptions are correct, and whether politicians actually allocate funds based on projected efficacy, are another matter entirely.