Economies of scale, usually. I've studied railroading economics a lot, and you ALWAYS go with the economies of scale. ALWAYS. They're spectacularly important.Here's a question for those proposing a second train over a popular route...
Would it be better to run the second train on the same route, with the benefit of more flexibility for intermediate point passengers, as well as 'economy of scale' using the same stations and other facilities; or would it be better to run a new route, tapping into new intermediate point markets?
Because of economies of scale, adding a second train per day to the same route probably *reduces* the budget you have to ask for from Congress. Asking a second route probably *increases* the budget you have to ask for from Congress. Which should you do first? The one which gives you increased ridership and lower costs, obviously. After you've done that, you have more political power and more money to work with, and THEN you can start talking about added routes.
There might be an exception: If there's a shortage of passenger demand on one route, and you really can't fill a second train, THEN consider a second route. For instance, I seriously doubt the level of total demand on the "High Line" through North Dakota and Montana. Running the second train on the higher-population route through Bismarck and Missoula makes sense. This is akin to the reason why the Silver Star and Silver Meteor follow the same route *most* of the way but take two different routes through the lower-ridership sections of North Carolina and South Carolina.
Another example of the same exception, because the current route of the CZ is a bad compromise: new CZ service should go via Wyoming. Green River - Provo service is pretty questionable in terms of ridership demand, though I wouldn't support cutting it entirely. Denver-Salt Lake is actually faster via Wyoming *and* reaches more population centers. The current route is mostly there to serve the ski areas, and they don't get much business from the west side. Adding additional frequencies should probably come in the form of separate Denver - ski area trains, and Denver-Salt Lake service via Wyoming. (We can debate whether to go via Greeley, or via Fort Collins, which is slower but has even more ridership.)
Both of these are cases where the current route is in some sense "wrong" and should be the secondary route. I wouldn't support cutting it, but future investment should go to the correct route. Another case is the Southwest Chief: added service must be via Wichita and Amarillo, which should be the primary route anyway.
In the case of the BNSF route across Iowa towards Denver, I would actually support cutting it in favor of what is now the Iowa Interstate Route -- if Iowa ever got serious about passenger service. The current route is minimal in terms of riders served. Anyone using the current route can drive to Council Bluffs, Des Moines, and Iowa City, which have far more online population -- *and* the Iowa Interstate route is more direct.
Two trains a day should take the same route in both cases. Take the economies of scale. Use the highest population route.For example, a second daily Chicago<>Denver train...same route as current, or via UP? Or, second Chicago<>St. Paul, via current or via BNSF?
Even when there are two or three high-population routes, such as from Cleveland to Chicago, it's better to go with one route to start with because you can concentrate track upgrades on that route. That allows you to speed up the trains on that route, giving shorter journey times. That increases ridership and revenue again.... once you have 12 trains a day running at high speed on the "primary" route from Cleveland to Chicago, then it's time to think about adding secondary routes.
Think about it this way: did it make more sense to beef up and speed up the NEC, or to instead run additional trains on the Inland Route, the CNJ, the Reading, the B&O, the Erie, etc? I think the answer is obvious. One way gets you a high-speed corridor, the other gets two low-speed corridors at twice the price.