I don't claim to have it all down. I view the game mainly from a strategy viewpoint. There is always a new technique or different timing to try. The complicated strategies aren't always better than the simpler ones either. Adapting the strategies I know work to the map I am playing is great fun. I also tend to look for the quickest methods, so don't tend to focus on passengers very often (in the 1800s there are plenty of exceptions) as I find that a good system takes some years to build up and doubling tracking is often required to run the extra trains. BTW, good reminder on the revenue increase available for the express trains. I didn't realize there was that much difference. (I will have to look for it, but I think the total difference may be only 60%, which is still a big difference.) One of the best tricks I know for rails is to study the map overlays to find where the economy is the strongest, with a focus on hauling processed cargo (Goods, Meat, Cheese, . . . etc.). Processed cargoes especially Meat and Cheese develop a 20-30k price difference per load pretty quickly giving a nice profit.
The guide you are thinking of is most likely "A Handbook to RT3" by Oilcan available about halfway down
this page. It is an excellent resource. It goes into company and train management and industry investment very well, but plays the financial aspect safe. I often make more risky financial plays. On the other hand, I view it as cheating to try to "steal" a cargo stream from an existing industry. I may cut off streams that are going right across the map, but if there is one Lumber Mill with 5 logging camps that are close enough the camps are turning a profit, I will never plonk a competing Lumber Mill down in a slightly better spot. It would give me more profit, but this seems so low to me. It can be argued that trains do a similar thing ("stealing" cargo from one factory for another), but I hardly ever buy single input-single output (simple) industries that don't get a natural supply of resources without a rail connection.
I agree Alternate USA isn't really that difficult. Last play, I only had to monitor Corn, the rest of the cargoes took care of themselves as I had a single city where I built enough industry to consume all the cargo I needed to haul in (example 4 upgraded breweries on one square in one of the cities in Texahoma). Having a simple strategy makes this one fun, otherwise it can be a hair-pulling experience. Shattered Union is based on this map, is easy, but has an interesting premise to get rid of the AI. I got lucky on this one and completed it in 4 years. Might be fun to try it when you finish Alternate USA. Will try to think of some of the harder maps. Not many players like to play in time-attack mode, so the hardest maps generally turn out to be medium difficulty. Most maps I get the goal in 1/2 to 2/3 of the time given, and play against my best time instead of to get the goal in time for the medal. There are a couple others I can think of for 1.06: Chile, Michigan1830 (like Panama Canal, there is a Platinum medal available for the best players), and Denver & Western. Trainmaster is also a challenge I would put on your horizon for one day. There aren't many maps, but game-play is more complex and challenging. Hawk's tip is good advice, many of us have 3 separate installations (1.05, 1.06, and TM). Reason: they aren't fully compatible.
How to find the barracks? Go to the map overlay for Weapons or Ammo. In this scenario some warehouses demand both cargoes, otherwise green triangles will be barracks or military depots. Smaller triangles will be barracks and larger ones military depots. It becomes pretty easy to find them with a little practice.
1. Yes, I got the goal. Random trains aren't the best idea. They will most likely run at a loss. Just a few running between the cities with barracks will work best for weapons and ammo. Troops do like to travel to cities without other barracks, so an extra train or two running general service from the city with barracks will help this goal. You don't need heaps extra.
2. Yes. The idea of using the custom consist is to make sure any troops available get on the train instead of sitting in the station while other cargo that is more profitable is hauled before they are. They may disappear while waiting. I wouldn't put them on wait-to-fill. Think of this train as a 5-car train, with the potential to run with 3 loads of troops whenever they are available. This way the train is making money, and all the troops are still getting hauled. Every now and then, a quick check to see where troops want to go will keep this running smoothly. Sometimes the troops may want to travel back to the barracks so it may not be a bad idea to run maybe 2 coming back as well.
3. Train density depends on cargo density. If using custom consists (efficient) maybe 1 or 2 extra trains than normal. With maybe 2 running between each barracks (substantial distance, 1 for close). Remember that troops are like passengers, they build up over the years. The trains between the barracks only need to be bought once prices at one of the barracks are orange from oversupply. So, don't buy them all at once. Add them as time passes and traffic builds up.
4. You will realize pretty quickly where they are consumed by the map overlay. You are right, the general public has no demand for them. The idea of hauling between barracks is to help demand on the map, instead of directly adding to your haul count. Thing of it as a string. As price drops at one place, cargo is moved to the next destination and so on spreading across the map for consumption. This cuts out some long hauls that cost a pretty penny in fuel costs. Consumption is necessary to keep demand high. (I think I did do the long haul to Zanzibar and distributed from there.)
5. It was awhile since I noticed if troops are loaded on express trains. I think they are? Anyway, I wouldn't bother with this. I generally wait till the later stages of a map until I have a double-tracked line before trying to set up special express runs. Your reminder about the extra bonuses available from the speed record trains makes me want to set some up again though. There needs to be a certain amount of traffic to make this worthwhile for me. Other players are much more into express, so it isn't necessarily a bad thing.
6. I try to avoid it, but for some of the goals here it is necessary. When done it should mainly be for finished products. Resources are best processed at nearby industrial facilities (factories). I don't think I ever had more than double track here. A better solution would be a double-tracked more direct mainline which doesn't pass through the cities. Cities tend to be choke points.
7. Fair enough. Maybe more specific and a few less will work better. How many did you have?
8. On occasion. Normally this is a sign of too much management on my part so I will try to find an automated solution.
Glad you are having so much fun with the game! This turned into a pretty LOOOOOOOOOOOOONNNNNNNNNNNNNGGGGGGGGGGGGGGG post.
Hope you get some ideas from it.
Will try his African dream now, even though this already felt like a dream playing the Drums of war! 
I have completed this map in the past, but didn't give it a write up. I must say that the Poptop guys did a nice job of painting maps, I guess they were paid to do it. Only recently with the method of painting maps from satellite image have the fan made maps caught up to this standard. I wish the same could be said of Poptop's event scripting and seeding balance, but the re-dos of their maps such as this are very enjoyable maps to play. (I also recommend Gilded Age based on the Building to Buffalo map.) Some of the previous posters were talking about 1.06. That's confusing cause this is a 1.05 map. It could be played in 1.06 with an advantage, but that isn't what the creator intended.
I really appreciate it 
