the swahili coast has been muslim since far before omani occupation. Kilwa and Zanzibar were muslim far before then. Islam was a big part of Swahili culture, and their folk religion is far less influential in their history than islam, especially considering other african cultures.
That's basically what I meant, is that the incorporation of the Swahili into the Omani empire was a lot more natural than the Arabic takeover of the largely non-Islamic Maghreb.
I don’t think it’s either or, and bringing in Yemen in addition to Swahili would bring the best of both worlds: another arabic speaking civ as well as one of the Indian Ocean’s most influential trading empires
I think it probably is at this point for VI. We still have two very large gaps on the African map to fill and that's ignoring large regions of potential in North America, Central Eurasia, and Southeast Asia, not to mention players clamoring for Portugal, Byzantium, Italy, and Assyria/Babylon. I can see the devs devoting resources to one civ along the Swahili/Arabic coast; I am quite dubious that they will want to make a second civ when they haven't even tested how popular one civ might be.
These things appear to happen in stages. You get the "big" culture first (England, Egypt, Greece, Rome), then when players want more you add spin-offs in later titles (Scotland, Nubia, Macedonia, Byzantium). You get the large, space-filling cultural admixture blob first (Celts, Polynesia, Shoshone, Morocco), then once players are used to it you refine it into a more specific culture (Scotland, Maori, Comanche?, Numidia). It's a slow, years-long process of experimentation and feedback, and so far the only real feedback Firaxis has at all for that region is that Oman is a popular mod for V and happens to serve as a really good big/blob culture to start from, while Swahili and Yemen aren't even on the modders' radar (well except Suk's civ VI mod, which we can't know what effect a later mod had on the development process and imo Suk's mod highlights how difficult it is to come up with a strong Swahili design).
Point being, many of the new additions to the roster were very conservative, safe choices. Scotland, Canada, Australia, Norway, Gran Colombia, and Vietnam were all within the top few pages of most popular/subscribed V mods, as were Greece and India splits. Cree, Kongo, Hungary, and a Polynesia split were a tad further down but still fairly popular. Nubia is really just an Egypt spinoff so that was guaranteed to be reasonably popular (and was still a mod that was somewhat subscribed to). Georgia was a meme. I think the only new civ which was actually a risk was the Mapuche, and that was because a redditor historian approached the devs with a lot of material and made a case for them. But on the whole, the devs have been sticking to civs which were proven to be reasonably popular through the V mods, effectively six+ years of free, self-organized market research on what players wanted to see most in VI.
I'm not saying Swahili or Yemen wouldn't make great civs (although in both cases I think the research to flesh them out would be considerably more onerous for the devs than for Oman which just generally has more to work with). What I am saying, however, is that the devs seem to me to be almost religiously vetting new civs against the data that they have in an effort to maximize appeal and sales. As far as their product is concerned, their shortlist has always been limited, with very few exceptions, to popular V mods.
From their perspective, they have virtually no incentive to even consider Yemen--Oman or otherwise--when the playerbase never cared about Yemen enough in six years to create a mod for V. This doesn't prevent Yemen from standing a chance eventually, but barring becoming a meme or some specialized historian reaching out to Firaxis, Yemen still has potentially
years to go before it rises in the playerbase's consciousness to the point that the developers--at least under the apparent methodology of choosing VI civs--feel it is a safe, marketable choice. The edifice is too big and the method is too reliable to sacrifice for the sake of smaller cultural interest civs without data showing that this is what players will buy.
(to that end, I do somewhat fear for the Berbers, in that Berber mods were never very popular. Granted, players already had Morocco and maybe that was enough, and the devs still feel the need to put something Maghreb in to fill the Moroccan void. But the Berbers would be relying more on Moroccan popularity as substitute for V mod popularity and I'm not sure how far that goes. Also, the American southwest was pretty spotty on mod quality and popularity so our speculation in that area is also generally unsupported by V mod market research, although again I'm not sure how far "substitution" goes if they went with the Navajo or Apache over Comanche--it's native tribes especially like the Cree/Mapuche/Maori that I think the devs also lean a bit on tribal member population size in absence of clearer indicators.)