Good luck dividing Precolombian civilisations and Subsaharan Africa, two parts of the world we have very limited historical knowledge of, basing on two separate (and kinda eurocentric) cultural indicators. Go on, do a research, it will be fun
However my main problem with this idea is different.
It doesn't enhance the gameplay in meaningful way, because it concentrates not on 'what game needs' but on meta historical issues of ethno-linguistic divisions.
It is essentially 'double current cultural % mechanic and make it overcomplicated'. What is the point of introducing 2 instead of 1 indicator if one of them is going to be static and dependant on another anyway? In game you either get humanism or click 'culture convert' on a province. This change would make you either get humanism or click 'language convert' on a province, only this time this process will be less transparent and weirder.
Also do not underappreciate the amount of research needed to separate language and ethnic barriers in 15th century world. IMHO way too much effort for not enough gain.
You're right, this would be a huge amount of extra work for no real reason.