@Black_Elk:
it becomes a full mod and an entirely new game.
Here’s a radical suggestion which you might want to consider – either as a practical project, or just as a way to generate some out-of-the-box lateral thinking.
There are many aspects of the Global 1940 game (and of other A&A games) over which people have expressed dissatisfaction and for which they’ve proposed house rule alternatives. Some of these proposals have targeted just one or two specific game elements and have involved changes of a very modest nature (such as adjusting the turn order and nothing else). Other proposals have been more wide-ranging in scope and more intricate in nature. Generally speaking, however, all of these proposals have operated under an implicit restriction: trying to stick as closely as possible to the OOB game in order to keep things recognizable. It’s a valid design principle, with the advantage that it makes proposed changes acceptable to a broader base of players, but the downside is that it acts as a straightjacket: it severely limits what you can change.
So here’s an alternative you might want to think about: remove the self-imposed restiction of needing to come up with the same game. Rather than starting from the premise that the end result of the design process has to be a game which is essentially the same as G40, eliminate this predetermined end result from the equation. Instead, draw up an inventory of: 1) all the things about the game that you absolutely want to keep; 2) all the things that you’d be flexibily prepared to either keep or modify or discard, depending on what works best; and 3) all the things that you feel don’t work properly in the OOB rules and that you feel absolutely must be improved or replaced, whether by minor tinkering or total reconceptualization. Then try to come up with a redesigned version of the game that addresses all of your issues and which works well on its own terms, regardless of how closely it adheres to the OOB game or how far it deviates from it.
The thing to keep in mind here is that, in the redesign of any existing system, you can only make so many adjustments to the system while sticking to its original basic architecture. Beyond a certain point, the original architecture becomes a constaining factor rather than a help, and instead of an improved version you end up with something that’s even clunkier than the unsatisfactory version which you were trying to fix in the first place. So beyond a certain point, you have to throw out that constraining architecture, start with a clean slate, focus on the specific things you’re trying to achieve, find the solutions which achieve your aims, and then let those solutions dictate the eventual new architecture of the revised system.