Construction on The Alco Line Project has started. As work progresses I'll be posting progress photos. Let's start with the design and the strategy behind it.
The first step is becoming clear as to the layout's objectives. In this situation, the customer, "Bill", had fond memories of the region and wanted to be transported to a time and place. When he walks into the layout room he wants to feel like he's "there". He wants to watch lashups of RS-3's throwing smoke on the wide-open prairies of Southern Minnesota.
Identify the key characteristics that define a place or a region's "essence", what makes it what it is. With the Alco Line, its signature look is one of vast prairie panoramas and wide open spaces punctuated by small "oasis" of towns spread fairly far apart. The distance between locations is a key feature.
This view, facing east from Sleepy Eye, gives a sense of the vast open areas emblematic of this area of the country. You can't capture this look on a layout if you insert more towns than your space will support. Set the layout's start and end points, the prototype locations where trains start and end their run. Next comes the most important design decision, deciding how many towns will be between those start and end points. It's this step that many modelers really struggle with and the primary reason why most designs miss the mark of capturing a place's "look". They try to model too much of the prototype. In doing so, the result is an overly compressed, toylike, "model railroady" look. The signature feature of the Alco Line is one of great distances and openness. If too many towns are included, it slips into more of a metropolitan look as opposed to that of being in the prairie. The key decision Bill made was to only model four towns. Two of those are at the end points meaning there are only two small towns (Sleepy Eye and Mankato) along the entire run.
By limiting the number of towns to only four, enough space is left for one of the most important defining features, the prairie. As design elements go, these "negative space" scenic features were considered as important as the towns themselves. This runs contrary to how most model railroad designs comes together. 
Diagram showing the portion of The Alco Line being modeled. 
Now, by contrast, if the natural tendency to include too many towns isn't held in check, the prairies, (negative space) are squeezed out. In doing so, the entire look and feel of the region is lost. When designing a layout that goes from place to place (as opposed to a switching layout), it's the space between elements that is crucial. That being the case, those "empty spaces" aka "negative space" need to be given as much or more priority as towns and industries. In other words, the square footage dedicated for "space between places" should be treated with equal importance as the towns and rail elements themselves.
Be very clear on your operating style. Are you primarily a rail fan who wants to do some light switching along the way? Do you plan to have "formal" operating sessions with crews? If you'll be having formal sessions, how often, and what size will your crews be? With the Alco Line, Bill will be mostly running solo, with his grandkids, or perhaps one or two other people. He's clear on how the layout will be used and has been wise to not include track that isn't needed to accomplish his goals.
Staging isn't mandatory. Depending on your scheme, in many (perhaps most) cases you don't need staging, especially if you're a rail fan. In this case, a train starts at Tracy or Waseca, runs the line, and that's that. Staging takes up an enormous amount of space and it makes zero sense to include it if doing so doesn't satisfy a crucial operational objective.
Don't go crazy with your yard size. Yards en route take up a lot of room and can consume a layout if you don't put limits on them. Unless you really love yard classification, make them just large enough to make the point, and leave it at that. For a theme like the Alco Line project, a passing siding, maybe an arrival/departure track, and two or three ladder tracks are more than enough.
Operational reliability is paramount. There are lines on paper or a computer screen and there is operational reality. They aren't always the same. Design aspects that prove problematic operationally don't always jump out at you from a paper sketch. You learn through experience. "On paper" it would have been nice to put Waseca on the left wall closer to the door. Doing so, however, would have required curved yard tracks. Coupling and uncoupling cars on a curve can be frustrating. Curved yard ladders need to be laid absolutely perfectly in order to prevent derailments. Even then, cars are more prone to jump the rails. We decided it was best to move Waseca down the line a bit in order to be able to have straight yard tracks.
Bill was very clear on what he wanted the layout to do for him and equally clear on the design strategy needed to accomplish that. The most important decision he made was limiting the length of the prototype boundaries represented. Doing so freed up a lot of space for wide-open prairie running, the essence of The Alco Line.