Based on the math we know so far, as well as the early gameplay I saw in Quill's Cree LP, it seems like keeping cities loyal, in the absence of significant pressure from opposing civs, will be far too easy. A new city, even settled half a continent away during a dark age, begins with its loyalty at 100/100 and rising, so long as no enemy city is affecting it. Given that this new city will immediately begin growing and producing its own loyalty, the situation will only improve from there. I don't think that players should routinely lose cities to low loyalty, but I do think they should have to make at least a bit of effort to avoid this: an isolated city should require some investment to keep in line, even if (or maybe especially if) it isn't threatened by a hostile power. Players have an extensive toolbox for dealing with loyalty issues, from assigning governors, to choosing policies, to building monuments and securing excess amenities, to simply settling cities close together. In most cases though (again, before the addition of opposing city pressure, which I don't have a good feel for yet), they can get by without even opening this toolbox. I would argue that cities should begin with loyalty trending slightly downwards and that non-foreign sources of disloyalty in general should be increased. I don't think that this would cause many non-pressured cities to actually defect, but it would force players to at least consider the possibility, which I think would be an improvement in both gameplay and immersion.