As a rule, and as an (at least formally trained) historian, I think there is a difference between documenting an era or event, and including the vocabulary and phrasing of the participants through primary sources, and commenting on or analyzing that era or event and using the same vocabulary and phrasing in a way where the reader might reasonably think you are normalizing or accepting that discourse.
For instance, in writing about the American South and the era of slavery or Jim Crow, one will of necessity have to include primary source accounts that use certain language to describe and label the Black population, enslaved or free. One does not, however, have to use that same language in the analysis and exposition parts of the work, the ones that are in the author’s own voice. Indeed, unless you are going for some pretty advanced rhetorical and compositional techniques, it’s generally a very bad idea to do so.
Games are of course not historical scholarly articles. The line between the game developer’s voice and the voice of characters or in a way the setting itself is sometimes rather blurry. One has to decide for oneself whether the act of creating a game about a subject replicates the work of the historian, where you can equate the language of characters with primary sources, and thus give them license to speak in historically appropriate if currently unacceptable ways.
Even so, using problematic language in other parts of the game, parts that might correspond with say a historians analysis of some primary sources (in this case, incidental game text, descriptions clearly not emanating from in-game characters, that sort of thing) is, well, problematic, though context is probably everything.
In general, in scholarship or other creative stuff, it comes down to what is your ultimate goal. Does what you include or omit support your conscious intent? Is it necessary for the thing you are making to do its, well, thing? If so, do it and own it, but be prepared to fight :).