score:10
It makes sense that colonies belonging to different European nations would preferentially buy from traders of the same nation, since under the prevailing mercantilist legal regimes the latter would automatically (or most easily) have the right to trade with said colonies. (And would know the language, which helps.) And, of course, it also makes sense that when colonists moved and took their slaves with them, or sold their slaves, it would be more likely for the destination to be another colony of the same mother country.
The problem in trying to make a simple conclusion like "traders from European country X always bought slaves of ethnic group Y from locale Z" is that:
Different European countries operated in different areas of West Africa at different times;
There were probably variations in the ethnicity of slaves that were available to European slave traders at different times based on local historical developments. E.g., if Dahomey lost a war, there might be more Fon- and Ewe-speaking slaves for sale from the victors.
Some colonies imported slaves directly from Africa, while others did so indirectly (via transshipment points or regional markets in the Caribbean, for example).
As you note, colonies sometimes changed hands, which could have some effect on the slave trade.
I suspect it may be possible to derive more information about this, but in many cases the records have not been investigated properly. A couple of decades ago, a researcher (Gwendolyn Midlo Hall) looking into 18th and 19th Century records in Lousiana discovered extensive notes about the particular ethnic origins of slaves, which the local archivists were unaware of: https://tracingafricanroots.wordpress.com/2015/09/25/louisiana-most-african-diversity-within-the-united-states/ . A relevant quote:
the complex colonial history of Louisiana, featuring periods of French, Spanish and American rule, has contributed to a possibly greater degree of African diversity than found anywhere else in the USA. Each timeperiod was characterized by its own distinct ethnic/regional pattern of slave imports.
(The downside is that records like these are probably rarer or missing for the 16th and 17th Centuries, so earlier aspects of the slave trade may not be as recoverable.)
Gwendolyn Midlo Hall later wrote a book called Slavery and African Ethnicities in the Americas: Restoring the Links, which I suspect might be relevant. (I haven't read it, but it seems to have gotten lots of positive scholarly reviews.)
As for "In the rest of the Americas, either Afroamerican religions were completely eradicated, or they are predominantly influenced by Yoruba tradition":
There are "Vodou" religions in places like Cuba and Puerto Rico, though these may be Haitian imports.
Obeah, practiced in the British Caribbean (especially Jamaica), is variously attributed to the Igbo, the Akan, and the Ashanti.
Palo in Cuba is thought to be derived from Kongo culture; its liturgical language is a mixture of Spanish and Bantu. (So Cuba is perhaps somewhat similar to the Brazilian case, in that you can have multiple religions from different origins in the same country, though as far as I know Cuba was always in Spanish hands.)
And Kumina in Jamaica is also thought to be derived from Kongo culture, but via post-emancipation immigrants (rather than slaves) from the Congo region.