"Infinite City Spam" (ICS for short) is a strategy of rapid expansion, putting down cities\colonies everywhere in the early game, rather than following a more balanced approach.

In some games it's very effective, others not so much. When it is the optimal strategy, it usually doesn't make the game fun to play, as it limits strategic options and introduces too much micromanagement.

Various games tried to address this issue, for example:

Master of Orion - many planets have terrain which requires expensive technologies to settle (e.g. Controlled Toxic Environment)

Civilization 4 - new cities lose money, moreso the further away they are from the capital, and the more cities there are. Only once they grow and get some buildings, do they become a net benefit to the economy.

What happens in Elemental right now?

New cities can build studies and arcane labs for as long as the global population allows. Often, these buildings give the best benefit for the time spent building, so having six cities continually building studies is a lot better than having three cities do the same (as bigger cities don't produce buildings faster than smaller ones, research will increase at the speed corresponding to number of cities, not reflecting their size). Same for units.

Also, new cities means new citizens. Each city's prestige is independent, so having six cities with +1 prestige each, rather than three, means that six people join the kingdom each season, rather than three. Since each new person adds more money to the kingdom, ICS also allows a lot more gold to be generated in the kingdom.

Overall, this means that ICS is probably the best approach to the game right now.

So, what are the options to limit it, and introduce benefits to bigger cities?

1) Limit the number of buildings that a city can build at a given size, limiting the benefits small cities can provide. I've only played 1.1 where the population pool is global, so perhaps having population as a city-local resource is a better approach (or perhaps it had its own problems?). Alternatively, we could limit the number of buildings that can be built in a city - so, instead of building Arcane Labs until the cows come home, we can have (each building can only be built once per city, unless specified otherwise):

Hermit's Hut (4 seasons, 1 arcane research, city level 1 required)
Arcane Lab (8 seasons, 2 arcane research, city level 2 required)
Hedge Magic Shop (8 seasons, 1 arcane research, 1 gold, city level 2 required)
Mage Tower (10 seasons, 3 arcane research, city level 3 required)
Wizards Council (16 seasons, +100% arcane research, city level 4 required, one per kingdom)
Sorcery School (12 seasons, 5 arcane research, city level 4 required)

This places a limit on amount of research a single settlement can produce (also makes lost libraries\ancient temples more valuable!). Also note that higher-end buildings are more effective than lower-end ones.

2) Citizens are not taxed directly. Small settlements cost money. Cities should only start to break even at size 2, after some money-making buildings have been built. Distance from capital may play a part, too.

E.g. a new city is going to cost 5 gold, on average, and no building available at level 1 provides gold. Once it hits size 2, it can build the following (each building is unique per city):

Market (8 seasons, 3 gold, city level 2 required)
Hedge Magic Shop (8 seasons, 1 arcane research, 1 gold, city level 2 required)

With trade routes and specialization in guildar production (say, a +40% to gold), it could be making 3 * 1.4 = 4.2 gold. Once again, in this scenario, citizens do not produce gold at all, only buildings do.

Since various buildings come at different technologies, advancing the tech tree and building up older cities should help with horizontal expansion. Because gold production is limited, Gold Mines become more important.

When city grows to level 3 and as new technologies are researched, additional buildings are opened up:

Jeweller (12 seasons, 3 gold, merchants sell magic rings, city level 3 required)
Tailor (12 seasons, 3 gold, merchants sell magic robes\cloaks, city level 3 required)
Great Fair (16 seasons, +100% gold in city, one per kingdom, city level 4 required)

3) Larger cities are more food-efficient. Food is a hard limit on the cities in Elemental until late game when Gardens are available (and at that stage, the map should probably be filled out), and larger cities provide more people per each point of food. This is already somewhat the case (Villa is better than House, which is better than Hut), but it can be made more gradual, to provide better benefits at every city level.
 
The following housing can be built once per city unless noted otherwise (note that the housing building names here represent not so much the individual dwellings, but rather defensible housing locations that the populace can retreat to in case of monster attack, which suits the atmosphere better, I think):

Hut (4 seasons, houses 15 people, requires 1 food, city level 1 - 15 people\food point)
Town Hall (8 seasons, houses 40 people, requires 2 food, city level 2 - 20 people\food point)
Merchant Quarter (12 seasons, houses 100 people, requires 3 food, city level 3 - 33 people\food point)
Noble Quarter (12 seasons, houses 100 people, requires 3 food, city level 3 - 33 people\food point)
Enclave (16 seasons, houses 200 people, requires 4 food, city level 4, unlimited - 50 people\food point)

Once again, various buildings come at different technologies, and here advancing the tech tree will help vertical expansion. Refugee Camps are also made more important (though probably want to scale them down to 50 people to allow a level 3 city straight away, but keep level 4 cities with bigger bonuses out of easy reach)

4) Monsters attack larger kingdoms. Basically, intelligent monsters don't like civilization spreading throughout their lands, and dumb monsters just want to snack - which forces people to huddle together if they are to defend themselves (or wander around nomadically in small groups, hoping not to be noticed and eaten, until they join your kingdom).

Essentially, have new monster lairs be generated randomly in kingdom's territories, with their level based on the number of cities as well as game turn, with the frequency dependent on kingdom size, and with a preference to be on kingdom's edges where it borders wilderness. The lair level requirements mean that the player needs to research Quests to find and destroy them, and possibly also train\equip heroes (Warfare); if the lairs are left alone, they will periodically generate monsters that can go on rampage, pillaging the lands.

This gives the player's heroes more stuff to do inside the kingdom, makes the countryside more dangerous in response to civilization's attempts to tame it, and gives the early to mid-game a feel of a few scattered pockets of humanity battling against hordes of darkness.

---

Ideally, with these changes, the following can be achieved:

Cities level 1's and underdeveloped level 2's that do not grab a strategic location or resource are expensive endeavours that can be afforded only by rich empires with a solid core, in late game.

Early to mid game is a scattering of kingdoms separated by stretches of wilderness. With less than normal number of civilizations in the game, the situation will last even longer.

Only in the mid-to-late game are the last patches of wilderness eradicated, when new settlements can build gardens to make themselves food-sufficient.