The general feeling I get from the current universe is a lot of it is economically non-viable. I would define viability based on the total cost required to get a material to a planet where it’s needed. If Planet A needs SF, then viable fuel producing planets are those who can produce SF at a cost that is profitable on Planet A.
Each planet has a best price for all materials in the universe that will dictate what can be produced there. This price list changes with the universe development as more planets are colonized. I say price, but I really mean time, because both are interchangeable. If a planet has rich deposits, but is so far away that it costs too much to ship in food, then it’s not viable. It could become viable if a food producing planet develops closer, which would then reduce its food cost and make it viable. If there are no viable agricultural planet, then it will never become viable itself.
It’s more than possible that such unviable clusters are generated, like the south-western part of the universe. Some other clusters may be viable, but fall into a chicken or egg situation where it needs to be built up simultaneously in a coordinated effort, like Hubur. Ideally neither would exist and settlements would organically spread outward as material availability increases with each new base.
I attribute it to base setups being fixed. To extract a given amount of resources, we always need the same amount of prefabs, workers and consumables.
What I propose is to create different versions of the same buildings with different construction parameters.
In a metal rich, food poor cluster, buildings using more prefabs and less workers would allow colonization by shifting costs from food to metals. For example, a Rig could cost 36 BSE, 120 MCG, but employ 15 pioneers. 300% materials increase for 50% worker reduction.
In a fertile cluster, the opposite could be available. That rig would cost 6 BSE, 20 MCG, but employ 60 pioneers.
Planets can then be generated with higher variance, creating massive resource differences between them. Generating average clusters won’t be necessary anymore. Even starter regions could have that huge variance to further put emphasis on their intended role and create trading potential.
These buildings could also modify area cost and worker requirements. That would be useful for new bases to initially occupy a lot more area to boost initial production and then shift to compact versions as the base fills up. It could also allow colonization if there’s really no metals or food nearby as a long term setup.
The optimal setup would always be the current building costs we have. These modifications would need to come with a penalty that makes them 100% unviable if the universe was compressed in a single point. They are only viable because they shift costs from a scarce remote material to an abundant local material temporarily while the cluster develops organically.