Out of 4 theoretical combinations of the City-Town system, only 2 are relevant on a closer look, so that the established Tall-Wide Dichotomy more or less stays the same.
But first I've to emphasise that in the new framework fully utilising your settlement cap is always beneficial. Because they're no workers anymore and border expansion happens automatically with growth events, Towns cost you only the initial Settler, while being practically micro-management free. They grant Resources, Gold for your Empire and Food for your Cities, while denying the land to your Enemies and creating a defensible Buffer to your core. (e.g. Walling a Town center on Hills in a Plain and stationing 1 Ranged Unit therein.)
With this in mind, the 4 theoretical combinations are:
I. Staying within the given settlement cap. Low share of Cities, high share of Towns.
II. Staying within the given settlement cap. High share of Cities, low share of Towns.
III. Expanding/going over the settlement cap. Low share of Cities, high share of Towns.
IV. Expanding/going over the settlement cap. High share of Cities, low share of Towns.
Despite Cities having more building options and more Production (Towns inefficiently converting it to Gold and having also only fewer purcheasing options) I and II are practically the same. Why? Well what is keeping you from converting all your Towns into Cities right away if they're so much better? Answer: 1. Strongly increasing Gold cost per conversion. 2. All Cities but your Old Capital get degraded to Towns in a Age transition, with all Building adjacencies being erased.
Now (1.) simply means that the marginal benefit of an additional City decreases. You're paying a large sum of Gold upfront to gain additional local production (and have more building options). This lends to the observation that Towns aren't that bad after all on a second look. Your local production is only worth something after you build sth. (let's say a yield-generating building), while in Towns you immediatly and constantly gain Gold, which you can store, use globally and with full flexibility. Also the Basic Growth-Focus of a Town (+50% growth) and many Specializations (like Farming/Fishing) are quite strong and cost-free bonuses. Furthermore the less production a Town has (e.g. being in a plains/coastal position) the less benefit generated from converting it to a City. Additionaly given restrains in land mass vs. your enemies you can efficiently and without much disadvantages settle marginal positions between/ on the edge of your current Settlements with Towns. (E.g. A coastal position with merely 2-3 land tiles for Quartes becoming a Food behemoth with +5 Fishing Boats for no cost.) So the main reason then for creating Cities is mostly the second aspect, to utilize a broader range of beneficial building options.
And speaking of buildings, this (2.) fact makes them be less worth than in previous iterations and as such also devalues the main benefit of more Cities, while in addition it makes the increasing Gold cost even more punishing because you've to re-upgrade a much greater number of Towns each Age. As a tall player you want to fortify your postion by focusing on Age-constant yields (or even Age-increasing yields), which are UQ/UI, Great Works, Specialists and Wonders. The importance of Wonders alone make the concentration of production in 1-2 Cities relatively more important then more total production across a higher number of Cities. (Yes on the other hand more Cities mean more UQ's. UI's can also be prucheased in Towns.)
Now with the addition of Specialists, which generate Science/Culture yields and copy 1/2 of present adjacency yields, there is introduced a further balancing between --> More Cities, More (of the same) yield-generating Buildings and --> Feeding Cities for Specialists. As the former option becomes marginally more unattractive, while the latter stays constant and is a perfect fit for a City-Town combo, we can expect to see in practice a fusion of I and II to a common tall strategy of having a continuum of reasonable City/Town shares, with the expected tendency of increasing City shares probabably canceled out by the automatic increase of the settlement cap each Age (see further below). So
--> Tall: Staying within the given settlement cap. Roughly constant (/Increasing) share of Cities over the Ages.
Regarding III and IV, it is immediatly clear that IV is nonsense on account of the Happiness mechanic. Playing ultra-wide you want to 1.Keep your global Happiness above zero and 2. Keep your local Happiness in your key Cities above zero. It's pretty evident that (2.) can be far easier achieved in few selected great Cities, then in a larger and possibly expanding set of smaller Cities. (Especially keeping Age Transition in mind) For playing wide generally another facet comes into play: Better City Specialization. With more Towns you can cost-efficiently generate with ease vast amounts of Food (and Gold), so you don't need to build these building types in your few cities, but instead concentrate more on Science/Culture/Production and Specialists. So
--> (Ultra-)Wide: Expanding/Going over the settlement cap. Low share of Cities, high (and increasing) share of Towns over the Ages.
Wide: Expanding the settlement cap (/going over the settlement cap a bit). Roughly constant share of Cities over the Ages.
Now this on it's own puts the tall player in a difficult position against a "normal" wide player. The latter has the same options, but more benefits than the former. Here the very strong leader attributes come into play, with fantastic bonuses to Towns and Specialists respectivly (e.g. +15/30% to all Yields, Maintance reduction). Civ selection plays also a part, with most Civilizations already roughly divided into the tall/wide dichotomy, usually by their abilities and number of unique Civics with settlement cap increases.
E.g. Rome, Persia +2 | Egypt, Greece, Han, Maurya, Maya, Missipian +1 | Aksum, Khmer +0
But we've to keep in mind that the practical differences will probably be not so stark, as the settlement cap is reset to a new basis every Age! So e.g. Khmer having a cap of 5 at the end of Antiquity nevertheless translates to a starting cap of 8 for Majapahit at the start of the Exploration Age. ( Exploration Civics increasing this +2 for a end cap of 10, with no cap increasing unique civics for Majapahit) So these Civ-Specific caps affect mostly intra-Age Gameplay strategy, but can also be used for dynamic transitions. E.g. especially the combination of a tall Civ in Antiquity which "blossoms out" into a wide Exploration Civ, to exploit the DL bonuses to a maximum, seems interesting. (With a side view to the Leader Attribute trees Culture/Expansion) Well this got longer then planned...
