Thoughts?
Sounds good.
Let's face it - the ultra defensive positions that caused you to introduce warp relays are
also a result of player choice in setup - players choosing to play on maps that have defensive terrain and/or enabling the Zenith trader rather than choosing to play on more open maps with less defensive terrain or without the Zenith trader. The choice of map and starting position all by itself probably has a larger impact on the total difficulty than any other individual factor apart from the two AI difficulty settings.
Thus having the AI suffer all the problems that follows from defensive terrain and fortifications (having difficulty beating it and/or bypassing it) sounds, as the default option, much better than giving the AI an "I'll magically teleport over defensive terrain rather than trying to bypass it on the tactical map if I cannot beat it" mechanic. If players play on chokepoint maps, it is reasonable to assume that in many cases it is because they expect defenses to be worthwhile and have the most fun gaming experience when they are, whereas players playing on non-chokepoint maps are looking for another experience.
So:
- Counterattack posts off by default because most people think they suck (for different reasons). Enable them as plot or get the Retaliatory Counterattack AI player to enjoy them in the game.
- Warp Relays off by default. Enable them to make it harder to defend yourself.
That said, I still think swapping the Core turret and normal turret restrictions is an excellent idea. It feels so silly that e.g. the find of a single Spider V controller allows me to build the best Spider turrets in the game everywhere, while painstakingly paying for Spider turrets with knowledge gives me a fixed amount of weaker turrets that, unless they are concentrated in chokepoints, are of little relevance. Defense in depth and playing with less defensive terrain would be a much more attractive option if one could actually build more defenses on more planets, so long as one's energy economy could support it. (Though this might requiring balancing the cap size and energy costs. Especially the energy costs for HBC's could stand to go up considerably, doubled or tripled, if they could be deployed up to a fixed number on a per-system basis.)