That was really funny. "Do you like chocolate? Well, according to multiple studies, you don't apparently."
How can a study tell me what is comfortable for me? Are we going crazy? We are starting to treat studies like little gods which tell us what to do, and apparently, what to like.
So I think conclusion such as "people think they prefer it, but actually, information retention is much higher with white background" is quite applicable despite you thinking that's not true.
Unless you are an emigre from Betelgeuse or a particularly intelligent octopus, you have eyes that work in the same way as everyone else's eyes. The biology and physiology of the human eye are well understood. Chocolate is an excellent comparison. Whether you like chocolate is a matter of taste. But you don't get to claim that eating chocolate for breakfast, lunch, and dinner is optimal because you prefer it.
Seriously, eye doctors are always interested in how I am even able to see. Which I wasn't at some time, until that got fixed by a brilliant professor.
So, despite the studies, for me white over black is easier to read. Maybe my performance is bad by using white over black. But it would be even worse with black over white, as I would spend all the time squinting at the screen.