We of course cant be certain of anything from such a small sample of previous changes. It could be an internal rule within the development team or it could be random chance. I think Menzies point is sound IF it is an internal rule and at this point we have no evidence either way that it is or isn't but from what we can observe it could be. As a physicist Menzies would understand the basic scientific concept that the standard answer to most questions is - We don't know. Many rules of physics were later proved to be incorrect once more knowledge was gained. Often that knowledge can only be gained with technology unavailable to the people making the original rules. Until people knew what it was like outside Earths atmosphere many things were assumed and some assumptions were incorrect etc etc. In this case we can assume that it's an internal rule and we can infer from that assumption that Venice has been replaced but since we aren't in the Firaxis development team we cant know for sure if the assumption is correct or not (given the small sample of data).
Menzies list of the last 4 is virtually the same as mine based on similar logic. I looked at the things we know about the new expansion and previous behaviour design wise and it said to me - Africa, Nth America, Asia & wildcard. From that I know the Moors was the most logical currently excluded Civ from Africa, I think Sioux is the most iconic of the Nth American tribes and therefore the most likely replacement to Peublo, that Majapahit fits the trade/tourism aspects of the game and that modern Indonesia is extremely populous meaning it is the most likely from Asia. The last Civ is a pure wild card as it could fit any of the criteria and be a justified inclusion. Italy or Venice would fit as a wild card here because in general people didn't expect it given Romes inclusion so it is a bit like the Huns, Polynesia etc and the devs like throwing something like that out there, it works for trade/tourism/ideology and especially for things like Great Works. It fits the Scramble scenario etc etc. Much like Menzies it's probably not what I would have chosen but I can understand why it has if it is in. Those last 4 civs could just as easily be Berber, Comanche, Vietnam & Australia or Benin, CSA, Timurid & the West Indies etc etc. It's fun to speculate and it's fun to try to use some sort of logic to figure things out but without knowing the development teams internal logic and rules in the end we really cant be certain until we get more data to analyse.