So what you're saying is that having a fully depleted loot bar should reduce supply limit by a ton; maybe up to 80%? Because when loot bar is empty, the army has run out of supplies to loot. Having logistics trains would still be good IMO though; having a chain of occupied enemy and unoccupied owned provinces to the capital or the nearest fort (including unblocked sea tiles) should increase the rate of reinforcements from manpower. Not having these would make the better option resorting, as you mentioned, to locals (i.e. Mercenaries recruited in occupied provinces).
I thought this already occurred in that looting reduces supply.
Ideally, though most forts should fall before the loot bar empties if empty . Also Cavalry would need some rework as a siege well stocked with light cavalry had much greater area it could pillage to maintain sieges.
The historical situation allowing *easy* conquest of Aztec wouldn't work very well in the game with Borneo being 100% one nation or an Aztec that is literally the only remaining tag in Mesoamerica though. Picking up local mercenaries would be viable to an extent regardless, but not the same extent as having obviously hostile nations to each other in close proximity. From the perspective of playing as the native it just sucks right now. You get none of the politicking/pre-war contact/context, just a blind declaration out of nowhere from a nation you can't even contact.
I don't disagree, I have never suggested that the game actually have 200 men armies take over major empires ... I just don't want a wildly ahistorical logistics system that will tank the AI. Likewise I do not see how adding a bunch of remote province in interior Africa actually makes playing natives more enjoyable. Right now if you want to play an isolated African nation with your back against the wall you have Kongo.
Adding in the Great Lakes State and ahistorical avenues of attack just makes all these states play more and more like everything else.
You run into a real problem from a gameplay standpoint if you allow easy pickups of merc from distant nations. It would destroy the viability of the nations and still wouldn't resemble a plausible model most likely, especially in cases where we have no historical examples against which to compare (a truly united Mesoamerica for example, even culture converted in game terms etc).
Europeans won with mercs against much more powerful states than united Mesoamericans. Long term European defeats are pretty rare be it in China or elsewhere. I see no possible way that a state with deep draft oceanic transport, cannon, and mass horses is going to lose a long campaign there.
I find this doubtful, since territory in Africa changes hands so rarely after the early goings. The Europeans typically leave Kongo and Kanem Bornu untouched, and I've watched Portugal fail wars against Mutapa because the AI doesn't plan transports out well. Sure, the player could bait the Ottomans there, but as you already pointed out the Ottomans are perfectly willing to bleed out their manpower long before reaching that region anyway.
If the Ottomans cannot get military access to your homeland (because Poland and Austria hate them), they sometimes will march doom stacks through the Sunni corridors leaving the homeland open. This will not get better with more states.
I'd like to see at least one more narrow corridor in Africa between West/East and a lot more vision for the sub-Saharan states in Africa (especially West Africa), mostly because of the limited diplomacy and constrained early game.
But this is literally what makes these starts unique. You cannot just go conquest in the other direction or create stable front with a game-long ally. Some starts should be limited diplomacy and constrained it makes them different.
I'm not sure why you think giving this region decent treatment would be such an issue for the AI. It's not a case where corridors between wasteland are even necessary. Only the deepest parts of the Congo rainforest and the Kalahari desert need to remain wasteland. And no, "being 80% surrounded by wasteland gives a different experience" isn't a good reason to keep Africa how it is.
Being surrounded on all sides by land that did not afford large scale state army invasions was the Kongolese strategic situation. If you look at their historical troop arrangements, their fortifications, their wars ... it all is oriented toward the Atlantic. You will obliterate all this history. If you want more states because you like states more than historical accuracy, that's fine, but adding in the Great Lakes as an East-West vector is by far the biggest deviation from historical strategy and conditions you could make in Africa.
Colonial nations are by no means a perfect abstraction (I hated it when Paradox announced them), and I would love to see a brand new system for colonization, but it's not practical. Currently the game handles the cases you mention with colonial nations and that would stay the same, since my solution is built around existing mechanics.
We have many recorded instances of the central European state directing fortification of key locations (e.g. Detroit and Mackinac) which in turn became settlements.
I frankly do not see what either of the proposals here actually seeks to add to gameplay.
Ahistorically opening up the Great Lakes region gives us more nations which will be so terribly fun because .... reasons?
Ahistorically limiting colonization will make the game more fun because ... reasons?
Ahistorically having a logistic system that discourages oceanic transport is more fun because reasons??
The most important historical things to maintain are the strategic balances around the globe. The best feeling of history is looking at the situation and feeling some of the same pulls that historical leaders felt. Throwing a lot of that away is a lot to pay just so we can have a few prettier blobs of color or such things.