Probably pointless to have this argument again, but I may as well post this.
You don't need to make a new name list for any of pre existing culture. That work was already done in CK2, so just reuse the old one. This new metod would be used exclusively for things that would otherwise get no melting pot cultures. It’s obviously pointless to substitute a culture for which the work is already done with a generic one.
When I say all Latin I very much mean all, which includes Iberian. A hypothetical melting pot Latin + Iberian may just recycle the Catalan namelist with some Occitan put into the mix. But when interacting with other cultures like (for example) Arabic or Slavic, the entire things gets threated as one source of names.
The trick is that you have two pools of names to pick from, so you can adapt either and end with something workable. Also you wouldn't just end with the same melting pot culture for each group, what we are recycling it's not the culture but the name list.
Let's use an example: Arabic + Slavic
We can form names for it by Arabizing Slavic names with some clever use of transliteration (Vladimir -> Fladimir) or by making Arab names sound more Slav (Ibrahim -> Ibragim). Bosnian is actually a great frame of reference as there is plenty of names of Arab origin. Do this process enough times and you end with a full Arab-Slav list of names.
But once you end with this list you don’t just make one culture, you make 3, each using the same name list. So for instance you could call them:
Arab + East Slav= Arabski
Arab + West Slav = Arabščina
Arab + South Slav = Alsulafia
After that you simply ignore any edge case. Could the East Africans met the Tibetans? Yes. Do we care about what happen in that edge case? No, because like 8 players are going to see it.
So the existing melting pots would be special exceptions to the "generic" melting pot code? OK, provided the exceptions work properly, and it doesn't become too complicated or start throwing problems.
Latin and Iberian are treated as different culture groups in CKII for several reasons, one of which appears to be to prevent one massive culture group from taking up most of western Europe.
But considering some of the cultures in there, they make no sense to be reduced down to a single namelist. If one of the parent cultures is Basque, it's going to contribute very different names to if the parent is Roman. A Basque/Dalmatian merger isn't going to look or sound like a Catalan/Occitan one, and the names from the Catalan list are largely inappropriate for that mix.
The three different Slavic groups are going to contribute different base names to the Arab/Slav mixed list, and again you risk inappropriate names being used for the namelist of the merger that are not derived from either culture that has been used for the merger, or that are clearly derivative of other, uninvolved cultures.
It's an interesting idea, but you're flattening it too far. Even within a culture group namelists can be very distinct, with names either changing to entirely different forms or being non-existent on one list, but very common in the other culture. Flattening multiple culture groups together strikes me as a bad way to handle it, especially considering the wide range of locations that a given culture group can cover before you consider that you're merging 3 culture groups together for creating the name lists. There's also the problem that you're effectively ignoring the differences between those cultures when doing this.
And then there's the issue that (say) a French/Lombard mixed culture might arise, and is then using the same culture/name list as a Norse/Roman culture. In these cases the ruling class and "peasant" classes are reversed in which culture groups they come from, but you're getting the same name list anyway. It just doesn't feel right; it feels too over simplified, and it feels like you'll end up with way too many names that are in forms foreign to both founding cultures.
And then there are the, as you say, "edge" cases. Where do you draw the line at what is an edge case? Remember that since any religion can pick up crusade-like traits, essentially any culture that is successful could end up "crusading" for land anywhere on the map, so what were unlikely "edge cases" before become more likely to happen. Your example with the East Africans and the Tibetans could, for example come about with both trying to "crusade" western India. They're more likely to come into contact in some ways than North Germanic/Arabic. The East Africans can get access by going around the Arabic peninsular.
A strong player of *any* culture can in theory take land of any other culture (although I'll admit some are less likely to happen than others), but I have, as a relatively poor player, ended up as an Irish ruler with a border to India thanks to a couple of crusades for the Holy Land and some NPC Holy Wars through modern Iran that I then inherited the profits of. It didn't last, because tanistry is awkward over those distances, but that's setting up Celtic/Indo-Aryan melting pots as a potential thing.
Not doing these "edge" cases also makes the "universal" melting pot idea seem very unfinished, and as if the cultures you don't give melting pots to don't matter, and are somehow not worth considering at all. I realise that's not your intent, but it has the potential to look bad.
In addition to this, a culture is more than just the name list. You need to decide on relevant special cultural troops, any special rules the culture might have (can they raid, can they access special successsions (tanistry, tibetan monasticism), do they get access to any special decisions (bear/stag/tiger hunting), do they have any special rules about being blocked from succession (castration/blinding etc.) and so on. You need to have a sensible rule for which culture group the blended culture goes into - and you have to bear in mind that even with the simplified version you suggest, you can have a Latin/Germanic merger and a Germanic/Latin merger that leads to (theoretically) the same blended culture in both cases, but the original ruling and "peasant" cultures are reversed, potentially affecting which culture group they should be in.
