It seems to me that the main arguments against snaking are about gameplay/strategy (i.e. teleportation, the feasibility of pathfinding, and game performance), while the main arguments for snaking are aesthetic. IIRC, games like Civilization IV and V would have the city graphic 'spill over' into surrounding tiles without actually expanding the city tile. Couldn't we try something like that with Elemental?
I was thinking the same thing.
First and foremost, because to me it is the most noticeable. Where 1-tile cities make the world look huge, snaking makes it look small. Suddenly I'm no longer the immortal sovereign leading my nation to become the dominant empire in the world, but I'm the city zoning councillor, having to decide where each and every building is placed, for each and every city that I build or conquer. It turns forests into parks, mountains into hills, lakes into ponds, and continents into islands. It just changes the whole atmosphere of the game for me.
Very good point, totally agree.
People will destroy buildings to make late game ones. Don't want that to be part of the game.
As far as the concern for destroying early improvements in favor of late game improvements, why not let improvements level as well? It could be a linear progression (lvl 1, 2, ... n) where the higher levels are tied to tech. It could also be a branching scheme:
Whipping Post: basic unrest reducing improvement
General Store: basic gildar producing improvement
Schoolhouse: basic research producing improvement
And so on, for each of the basic effects improvements offer.
Now, certain techs will permit these basic improvements to merely level up; they'll provide more of whatever benefit they've always provided. Certain techs will also permit improvements to "branch": a level three general store can become a level one market square or a level one traders guild. Each branch greatly enhances whatever benefit its root improvement provided, and perhaps some secondary benefit, as well as a penalty on some other aspect of the settlement.
Coupled with a limit on the number of improvements a settlement can have (even if this limit can be increased/decreased through gameplay), this would increase the strategic importance of improvements.
Just my two cents for one tile settlements.