You'll see wandering settlers quite often in the mid- to late-game, presumably produced because an AI realises it needs more cities to catch up with the competition, but left unsettled because there isn't actually a decent location in which to build. As much as the new hard-coded personality traits have improved the credibility of the AI, by giving them distinct approaches to diplomacy and guiding their actions to an extent, the basic triggers that decide what should be built and when sometimes seem to get stuck in a loop. When an AI civ follows one of the former types, there's a tendency to produce avalanches of apostles, sometimes long after every city within reach has been converted, and there is a recurrent problem with this kind of behaviour.
Broadly speaking, religions fall into one of two camps: they're either active in the world, taking strength from conversion, or self-contained, bringing great benefits to home-grown followers but tending not to proselytise.
Later, you're able to define more traits, which allow for production of buildings with various properties, or affect the way your faith changes your cities and units or those of your neighbours. As in Civ V, upon founding a religion you choose a pantheon, and many of these give boosts based on terrain and resources. Religion, particularly before it is properly organised, is more intimately related to the land. In the late stages of the game, there are so many policies to pick from, and so many slots available for cards in the modern governments, that you can push your people in just about any direction. I've deliberately begun periods of reconstruction and actual artistic renaissances or ages of scientific enlightenment, purely by altering government to suit the mood of the times. The risk is that if your initial assault isn't strong enough, you might be left in need of more units without the boosts used to build the initial force. That allows for a government to suit just about any scenario and makes for much more interesting choices and stages of play – you could gamble by switching to policies that boost military production before entering a war, for example, and then switch out those policies for ones that keep the folks back home happy once the armies are built. There are still base governments to choose from, unlocked in groups of three as you move through eras, but they're modified using cards unlocked on the cultural tech tree. Government and religion are both customisable, the latter in a similar approach to Civ V, the former in an entirely new card-based system. If you're chasing religious advancement, for example, that has an impact on the other cultural policies you discover and unlock, and certain relationships will be affected by your application of religion. This is a very busy Civ game, with lots of different cultural, technological and military plates to spin, but nothing works in isolation. As it is, the genius of the game's design isn't in the ideas themselves, but in the careful stitching together of those ideas. If this new engagement with the map had been attached to just one feature, it might come across as a gimmick, or the foundation for a concept not yet fully constructed. In Civ VI, the land makes a mark on you, just as you make your mark on the land. Few 4X games emphasise the importance of geography to this degree.
What I can say is that it's a radical redesign, true to the spirit of the series but finding new ideas in its elevation of the map from backdrop and resource-container to new plane of strategic and tactical importance. The fourth game has had ten years to work its way into my mind and it has endured, Civ VI is still young. The question I've been trying to answer, as I spent a final few hours with Civilization VI before writing this review, is simple: is it the best game in the series, or the best game in the series bar Civ IV? After more than a hundred hours of play, I still can't say for sure.