We've done this a lot, with the same goal as you: to explore historical settings. In my experience the problem with doing this (and also a main benefit) is that people don't make the same ideological shortcuts as normal. When we play medieval fantasy or sci-fi or whatever new people tend to apply all of the racist, sexist, anti-religious, ethnocentric baggage they have in real-life to the game even more than normal because the game feels like a safe, non-judgemental space and so they aren't worried about being attacked for their unpopular beliefs. On the plus side, this makes discussing these beliefs via the game easier in some ways, because people don't start off defensive, but it also makes it harder because people don't start off critically engaged.
When you use a real-life historical setting with which they are unfamiliar, the default reaction is diametric. People immediately worry that their lack of knowledge will brand them bigoted in some way, and their shame cripples them from being willing to engage in any but the most cautious of creative expressions. When we began discussing this behavior with people, most people (especially the less self-aware) would indicate that they felt uncomfortable because the setting was unfamiliar. And sometimes resolving this would resolve the problems-- if the GM in collaboration with the other players talked to the new player about the historical setting for a couple hours or so, the new player would in fact then be comfortable with the setting and the game could be played. Further investigation, however, revealed that the purported lack of knowledge was not in actuality the issue. Sometimes we would play games where two different players were both independently familiar with the setting from different backgrounds. Other times, someone would read up on the setting to solve the 'feeling uncomfortable' problem. In neither case did things go well. In the first case, the game would usually result in a very impassioned argument between the two players in question, with strongly-worded personal attacks, and the game would have to stop and it would take days or weeks before the people in question were on good terms again. In the latter case, the person would begin the game not feeling as comfortable as they expected but expecting it to improve, which it usually did, until something they liked in their reading conflicted with what the other knowledgeable person thought and then they would either get upset and shut down and stop participating or get extremely anxious and stop creatively participating to instead try to guess what the group thinks a good person would have their character do with each action.
Our core group didn't generally have this problem, because we have long dealt with extreme ideological differences between players in regular play as well and so we've largely become comfortable with calling each other terrible people via RPGs. When playing with people outside the core group, awareness of the problem has seemed to help a lot in getting things to work out well, as has being explicit about our 'The GM's ontological and political beliefs are true in the context of the game world, but are open for discussion via narrative analysis' social convention.
If you don't already, having some agreement about who gets to decide how people work and what it means to be a person and all of that stuff is a critical first step in being able to play games like these without getting into giant fights over real-life issues in disguise.