Defences are cheap and quick to build, but they have higher maintenance cost than building cost (it costs more to maintain than to build, so it's not a one shot investment for security), so you only maintain them when you need them. They would also be very hard to destroy by military units (unlike ships). They would also have a big defence potential, with a high hard maximum (maintenance cost sets the "soft" maximum) and a much bigger defence_power/ressource_spent than ships.
I think planets should generally be vulnerable targets, requiring fleets and/or expensive and time-consuming (to build *and* maintain) orbital defences to defend. If a planet can quickly build powerful defences, then there's no point in trying to outmanouvre an enemy fleet to take out a specific powerhouse production / research planet behind the front lines. Similarly, if you can build up and dispose of your defences on planets quickly and easily with no significant cost (since it's only maintainance that's expensive), then you don't even need a defensive fleet. You could just sit and build and research, confident in your ability to fend of any attack that comes your way by rush building massive planetary defences...
I don't like autobuild of different types of building ... I want to have the choice to pay for maintenance or not, not having a hard limitation on building quantity (unless they are unique by design, like a Civ Great wonder).
There are two kinds of "buildings", I believe:
1) The Civ Great wonder kind, that are quite a large undertaking with big benefits. You would likely specify when/where to build one of these.
2) The generic factories / farms / mines / labs, which are built automatically according to the focus setting of the planet. There are dozens or hundreds of these on a planet, simply indicated by a number of each. The details of cost, maintainance, upgrading, when to build, damage, repair and such probably haven't yet been determined... but it won't be direct player intervetion / micromanagement, I imagine. It would probably almost always be beneficial to build more of these, though environmental damage or increasing upkeep could be limiting factors.
Discussing "autobuild" needs to specifiy which of these is being discussed...
The latter almost need to be automatically built, to avoid huge amounts of micro which don't add much to the game. These probably can't be easily abstracted away however, without leaving the player confused as to why the production of a world is what it is, why changing it is taking as long as it is and other such things. The emergent dynamics of a system as described for farms / mines etc. would likely be more interesting and intricate for gameplay purposes as well... (as opposed to any abstract made up delays or equations to follow when dealing with focus effects and changes).
I like minefields, that's the way Stars! made space non-uniform. I don't know the consequences when mixed with starlanes.
Minefields could be located in a particular system only. They would be a factor in tactital combat... not in interstellar fleet movements.
I could go either way though... no strong opinion.
What's the good to have 120 planet when you just build things on 20% of them? Just have a galaxy with less planets.
Why have many planets per system if there exist only system wide defence and no planetary defences? Just have it abstracted as one "planetary system".
You claim to be all for great gameplay, yet you sacrifice ideas for the sake of clinging to the Master of Orion models.
More stuff = better epic sense of scale
Many of the conventions (multiple planets / system, dozens to hundreds of systems in an empire) aren't specific to Master of Orion, or even the 4X genre... they're the popular conception of space opera in most people's minds. The Dune series or Star Wars or the Foundation series or Trek all have single political groups or empires covering many many stars (as do the 4X games).
I'm sorry, but if you want to do a Master of Orion game, there will be a complexity from the scope of the game world. Those complexities are bound to create a lot of management. The key is to have tools to make that management macro management (high level decisions) instead of micro management (low level decisions).