I like the Quarter to Three concept which is a lot like the Civ4 concept. However, I'd tweak that to be a bit like the Civ2 concept where each region has natural resources, but there's no need for the micromanagement of say building a mine on the iron node. I'd sum it as, Complexity rocks, but tiles should be looked at and not fiddled with. (edit: basically a tip of the hat to Paradox's Victoria system)
Rather it'd be that most mountains have iron, so if your city is in a region with mountains, then it has access to irons. Trade routes then share the resources with other cities; Like in Victoria, trade routes can have their own efficiencies (regional infrastructure). Cities in the trade network, and having key production buildings can then turn the resource into something useful (e.g. forge turns iron into weapon, or building supplies).
There can still be some variety to the implied resources. E.g. some mountain ranges might tend to have copper, not iron, some plains would tend to have buffalo, not wheat. So there's still incentive to expand to different parts of the map, but less checking to see if a mine has been built for every type of mineral, or a farm for every type of food. The exact resources grabbed would still be by the cultural sphere of the city (i.e. tile-based grabbing of resources, rather than pre-defined regions).
Exceptions to the rule of no-tile fiddling would be rare items might be represented by nodes that need physical infranstrure. E.g. Admantine node needs a mine that connects to a city to supply that rare item to the trade network. E.g. the fire mana vs. the death mana for MoM.
So in other words, everything is a resource, and any Victoria or Settlers-like (remember that old Blue Byte/SSI/UBI game that keeps resurecting?) chaining of resources to prodoucts to products can be envisioned, but there's no need to micromanage the map to be sure that you workers fiddling to build tile upgrades (there can be an actual flow-chart screen that dictates, say 50% of iron made goes to the weaponsmith, 50% goes to making nails). If regional resource gathering upgrading is needed, it should be done at province/city level, not at the tile level.
That should reduce some useless CPU and player fiddling. To starve a city of resources, a proper siege should be conducted, not tile fiddling of military units, except in certain situations (e.g. horsemen pillaging the countryside abstractly).
Or in other words, Settlers/Victoria like economic complexity, Civ4 cultural grabbing of resources, Civ2 regional resource dispositions, Victoria-like regional efficiencies to resource gathering efficiency, very little tile management (very un-Civ4 like in that regards).
Sounds cool to me. Requires thought from the player to devise an economy on multiple levels, yet eliminates tile micromanagement which just sucks up CPU cycles and is just busy work for the player. Limits tile managements to truly rare resources by which the game might be lost/won, such that guarding them with a giant castle and army is really useful.
Afterthought: Theres still the issue of quantifying the actual resource stockpiles of a civ, if that's still in, but that can still be done fairly abstractly at the region level. E.g. Theoretical maximum resource gathering rate based on surroundings, influenced by the efficiency of gathering for the region/city, and further influenced by whatever workforce is assigned to the gathering.