Caps aren't going to change the balance in empire size between the two options. In order to have a place for an econ station, you need to have several worlds around it either neutered, or controlled, making the econ station the defacto large-empire solution. Even though it has a cap. You can't even HIT the cap (for MkII+III) until your empire exceeds some 15 planets, depending on layout (minimum would be 14, if you have an end on a snake map: home + 12 econ + 1 border world)
Why is it necessary to have the border world if I'm playing in a snake (or X, or other easily chokepointed map type)? If I'm going to make a fortress world, the command station type that I use doesn't really matter, because:
1. It's going to be on the wormhole leading further into my space, or otherwise far from the wormhole(s) enemies will enter the system through.
2. It's going to be buried under as many forcefields as I can throw on it while still blocking the wormhole(s) leading into my space to a reasonable extent and leaving a few to cover the home command station.
3. I'm going to have at least half, and probably more, of my turrets (and any other defensive structures I have) on that planet.
4. It will have a major minefield around any entry wormholes, and some strategically placed tractor beams and gravity turrets to hold enemies where I want them to be.
At that point, the type of command station is immaterial for all but the heaviest attacks or waves that ignore tractor beams and gravity effects, and the firepower I provide for defense should be able to handle that kind of attack on a fortress world. Certainly the military command stations are slightly superior in terms of survivability, and logistics stations increase the amount of time my turrets and other defenses have to shoot at AI vessels, but if an attack can overwhelm my fortress world having a non-economic command station isn't going to greatly increase the ability of the fortress world to hold out, and it's not like there's going to be anything behind the fortress world to stop the attack until it hits the homeworld (with the exception of Fallen Spire games).
I'm also not convinced that the economy needs to be significantly readjusted. I can have a ridiculously strong economy if I spend ridiculous amounts of knowledge to get there, and that's fine. I can also have a very strong military for about the same amount of knowledge, and play with a weaker economy. Neither way is particularly better than the other. From what other people have said, the expected number of planets that would be taken is somewhere around ten or fifteen, so you're expected to have 40,000 to 55,000 knowledge to work with in any given game, plus anything you get from knowledge hacking or destroying neighboring worlds to collect the knowledge. If you go all-out for the economy, you're going to spend 18,000 of your knowledge on things that cannot attack enemy worlds or defend your worlds, and that represents half to a third of the knowledge you are expected to have (assuming that the ten to fifteen worlds number that someone mentioned earlier is accurate). You're also going to limit the options you have when placing command stations, because if you're going to spend 9,000 knowledge unlocking economic command stations you really should use them.
I'd also like to point out that six Mark II economic stations is equivalent to about four or five worlds with an average of four Mark III harvesters each, in terms of resource income, not including any income from harvesters on the worlds that you put the Mark II Economic stations on. It's also equivalent to about eight worlds averaging four Mark II harvesters. Six Mark III economic stations is about equivalent to nine worlds averaging four Mark III harvesters in resource income. I don't see this as a particularly bad trade-off in terms of knowledge investment.
Edited to correct a math error.
