As noted, I completed the UHV (Monarch/Normal/600 AD, on latest Git). Here's a quick write-up.
Basically adopted the strategy I (and TJDowling) noted above. Tech wise, from the very start, you want to bee-line to Education. Among the other Muslim civs, you should be able to leverage Civil Service and Education for some techs.
After Timbuktu/Djenne have Markets/Aqueducts/Mints, I switch to Regulated Trade/Meritocracy, for stability and because they're probably the best civics you have in those columns (+1 Commerce on Workshops/Watermills and +1 Commerce on Specialists, respectively, for economic bonuses, and then the huge stability bonus from the combination of them). For the UHV 2/3 stretch, this means you'll be running Despotism/Meritocracy/Manoralism/Regulated Trade/Monasticism/Sovereignty. Once you get to Feudalism, switching to Tributaries is a good idea; Government civic is flexible - Monarch if you need stability, Despotism for infrastructure, Elective if things are going well as it's the only Government civic which boosts your econ. Notwithstanding the situation in some screencaps below, happiness shouldn't be an issue: Gold, Gems, Cotton, and Ivory should provide room for growth.
As noted above, Timbuktu and Djenne are the core cities, pumping out Sankore/the Prophet and Great Merchants respectively. Due to the Mali UHV, if you're running 100% gold working tiles is better for gold generation than running Merchants, so once Timbuktu is done with Sankore it drops specialists and just works the flood plains, with the improvement set-up observable in the screencaps - a mix of Workshops (so Timbuktu can build Sankore) and Cottages. Gao does nothing but pump out Workers and Settlers: since the Mali UHV applies to the tile a city works on itself, your cities actually pay for themselves almost immediately, which means founding a lot of cities which only work one tile as they'll always pay off. In particular, founding them around the minerals around Djenne frees Djenne up to work Merchants. City priority is: Timbuktu (Capital for happiness, culture, and maintenance) -> Djenne -> Gao -> Accra (to reach Crabs for health) -> North Nigerian city (to work Ivories/reach Cows for health) -> aforementioned garbage 1 pop cities around the Saharan Oases. Once Timbuktu and Djenne get their core infrastructure out, they spam Kebolos, Camels, and Archers so that you don't die to anything.
A lot of what's written here isn't necessary, but the result is that the game is won by around 1350 AD. This also isn't a very ambitious strategy - it just clears the UHV and doesn't necessarily get the maximum possible score. Here's the key checks:
First, my GM was generated in 1190, and reaches Mecca in 1260. It sits there as long as possible to maximize Mecca's pop growth and thus your GM gold yield. I would get +1800 with 6 pop Mecca ultimately in 1350. In retrospect, the GM could've also floated around to make contact with the Turks and Ottomans to have some more trade contacts. This turn is also the turn I finished Education, and I would crank down to 0% Science.
Second, this is a screencap of what the core looks like for the improvements. Generally, from the start of the game, the workers improve directly around Timbuktu, then sweep east to make the Saharan road in the top-right of the screencap, then sweeps back to the south of Djenne to improve everything. Also note in the screencap that Timbuktu has 7 turns before it generates its Great Prophet and finishes Sankore, which leaves almost 15 turns of flex room (Make sure to micro Djenne's specialists so that Timbuktu generates first). In one turn, Timbuktu could also rush Sankore with 4 pop if you have a super Arabia/Moors running around who could compete.
Third, this is a screencap in 1450 AD showing when UHV 1/2 were cleared. It's also observable that with 10 turns left until the first gold check that 5000 Gold has already been reached.
Lastly, here's a screencap in 1490, when the game is won. Doing the math, and assuming no growth and no decrease, I'm generating 175 gold per turn with 42 turns left until the second gold check (+7,350 Gold from slider generation), and a treasury of 6,774 gold, meaning just with slider I'll reach 14,124 by the check under these conditions. That's just one Great Merchant, and as seen at the top, Djenne is 11 turns away from an 88% (probably 70% by the generation due to the one Engineer) Great Merchant. The game is won at this point.
This is a screencap the turn after showing stability, which will only get better from here from improving the remaining land (notice on the bottom right how nothing's improved there) and from Cottage growth.
Mali's UHV is not a super challenging one once you figure things out since it's a puzzle UHV. From the above screencaps the only challenge left is staying alive, and there's no issues with that as you have no natural rivals who will conquer you, and even if you did, your cities have nothing to do but spam units. Padding score is possible through expansion and if you actually teched a bit. From the last screencap it's just an end turn exercise until 1700 AD, so I haven't bothered but it's obvious the game is won. Saves are added at regular intervals to show how thing went if someone cared enough to play it out.
Edit: Typos; added saves.
