Right, so I've been working on a scenario set between the two world wars (Oh, no, Virote, you're not working on another scenario, are you???), which will probably never see the light of day (due in part to the fact that it uses some of other peoples' ideas, and also because whilst I can map out the political landscape well, I'm not so big on the military landscape). In the scenario, the world is divided up between regions; America, Europe, Africa, Pacific (Australian and NZ) and Asia, with the world divided up in such a way that, for the most part, the resources won't intermix. (Apart from in Russia, where Asia and Europe intermix although I'm designing the European states so that they can only build European troops if said situation arises)

Asia here is the problematic one, particularly with regards to the Middle East. At current, from Korea and Manchuria in the North East to Palestine in the South West there exists a trade route, passing through Indochina and India, through Iran, and out to the Middle East. And since they all contain the same resources pretty much, it means that some units will be built in an area which they are out of place. So, do the French build Arabian-looking units (since they control Syria) or East Asian looking ones (since they control Indochina). In an ideal world, they'd be separate, but the problem exists with knowing where to draw the line. Here are the options as I see it:

1) Keep it how it is, perhaps giving the Middle East some Arab-spawning improvements for the French and British. This is imperfect, but from a tactical point of view, the regions are connected!

2) Divide the line between Iraq and Iran. This allows the British to project their influence into Iran. The downside of this is that Iran has nowhere left to run; they will inevitably have to attack the jewel in the crown at some point, and don't have much else they can build up on to build up their infrastructure (the only other civ in the region will be 3 Afghan cities, which in this scenario is controlled by the Arab states)
Iran will, of course, only be able to build Middle-Eastern looking units, but others, such as the French if they take it (which is very unlikely!) will be able to build Asians.

3) Divide the line between Iran/Afghanistan and the British. The biggest downside to this plan is that the British can't really project themselves into Iran or Afghanistan so much, and since both the British and the Russians are out of the picture, Iran's going to have a pretty easy time. The advantage is that with Afghanistan connected to the other Arab cities, it can still act as a counterbalance to Iran's power in the region, with the two pretty much being on par. This is probably the best route to go down for graphically, as well.

This is all I can think of. Perhaps there's another alternative which I haven't though of, I dunno. But anyway, what do you all think?

Afghanistan has always been a bugaboo for invaders - at least from Al the G. on. The period you're looking at had the tail end of the Great Game going on - Afghanistan & Southwest / Central Asia (including Tibet) being the prizes. So I'd favor option 3 - making it as difficult as possible for the imperial powers to establish hegemony.

As to the troop issues, I can think of two solutions. Use "city radius" resources as a requirement to auto-build "native" troops. Or use a "foreign legion" unit that would make sense in all those regions. In addition to the French both the Spanish & the Dutch had actual legions standing as military organizations separate from their regular armies during the interwar period. It wouldn't be too much of a reach to expand this for game purposes to include such things as the British colonial troops which still had a combination of local enlisted & British cadre. Also note as support of the latter that citizens of countries such as the US that were officially neutral enlisted in other nations' armies and often served in units that preserved their national identity.

Do not forget that the British can still recruit Gurkhas for their Gurkha battalions. Allowing the British to have some colonial units with British officer cadre makes good game sense.

Thanks for the answers. :)

For both the Gurkhas and the Foreign Legions, I'm thinking of having them auto-produced from buildings. However, I don't really want auto production to go further than that in the scenario, really.

Interesting point on Afghanistan. I had been swaying more towards option 3 beforehand, and had forgot about the Great Game...