After playing several games as the malians, and playing mutliple games against them, I feel that the choices between the landmarks have ALOT of room for differentiation.
The malians really are set up for a perfect choice between a food economy, or a gold economy.
The food economy is scout cavalry, backed by spearmen, archers, and javelin throwers.
The gold economy is mustofadi warriors, gunners, and sofas.
Now, look at the landmarks. The farimba garrison converts all the food costs to gold, and reduces it by 20%. On the other side, you have the cattle landmark (which gives lots of free food).
Then we get to the next choice. The castle that gives ALL units stealth (when it advertises only giving the mustofadi stealth which is a minor ui bug) vs the holidays for seige, unit spamming, or… FOOD generation.
Even in the first age, the choice is between gold and stone… or extra food from trade routes. So, the question is, what changes would emphasize this dichotomy even more?
Well, for starters, lets take a thought experiment. Lets say mustofadi warriors had their gold cost increased by 25, and their food cost decreased by 25. The resulting cost is 25 food, and 55 gold. Now, consider the math around taking the farimba garrison. The cost is (25 +55)*.8 = 64 gold.
Now lets say that the cost were exchanged, not 1:1, but for example 1.5:1. In the following example, the mustofadi warrior would have its gold cost increased by 10, and its food cost reduced by 15. The result is a food cost of 35 and gold cost of 40. This translates into a gold cost of 60 in the farimba garrison.
But lets say you went with 3:1. The mustofadi warrior has their food cost decreased by 30, and their gold cost increased by 10. This yields a warrior with a cost of 20 food, and 40 gold. In the farimba garrison, the cost is now 48 gold.
The most extreme of this occurs when the price of food for a unit in the farimba garrison slides below 20% of the cost of the gold cost for that unit. An example of this is a mustofadi warrior with a gold cost of 60, and a food cost of 10 food. If you put this in the farimba garrison, the resulting unit costs 56 gold. Which is free food, and less gold than making the unit normally.
This fully emphasize the “what could have been” for malians. What happens if the farimba garrison were to include a tech that increased the base gold cost of all units, but drastically reduced their food? This would dramatically alter the malian economy.
On the flip side, what if the extra food from cattle landmark had a technology that reduced the gold and wood costs of units, but also increased their food costs? Manipulating the costs of units in balance between the gold economy and the food economy leads to two dramatically different playstyles for the malian civilization. Even more dramatic than what we have now.
On the notes of the final landmarks, the stealth fortress gives non mustofadi units stealth. And that makes it less of a choice between the mustofadi and the low rank spear and javelin wielding fodder. A final tech from the stealth fortress that buffs the stealth duration of the mustofadi, or causes outposts and keeps to also grant the stealth aura would make the choice even more emphasized.