It would encourage blocky-shaped stations :p
Would it? I can't see someone planning out their station according to gravity modules. Personally, I would build my station then go back and min-max how many gravity modules I need to place along the structure I just finished.
A lot of chnucks. Either a lot of calculations or missing granularity, except if based on gradual changes between the settings from 4 nearest points when you are yourself inside a tetra (3D triangle).
[/QUOTE]
I'm not quite following what you are saying, so I'll clarify what I meant.
A planet spawns and takes up a 30x30x30 chunk area of physical blocks. We then assume another 10 chunk area for atmosphere. Gravity chunks are 5x5x5 areas, so we would need (8x8x8) 504 areas of gravity for this planet (though most of these areas will be pointing in the same direction). This isn't a big set of calculations, it's just a grid. The players position when near the planet is rounded to the gravity grid, then direction and force are applied according to that quadrant's values. Regardless of planet size, the calculation is the same. The file size for the planet will be a few kilobytes bigger, that's all.
As for atmospheric gravity, in my example planet there would be two layers of atmospheric gravity. The atmosphere is 10 chunks high, but gravity chunks are only 5 high. The outside most layer points down with strength 2. The inside layer points down with strength 5. When a player passes from one layer to the next they fall at a faster speed.
The gravity on face plates now looks like an upside-down pyramid. To fill in any gaps that might appear between the face plates, the gravity for that area can be angled to push the player into a nearby pyramid, then fall directly down. So it works similar to the water mechanics in Minecraft... only in 3D and with different speeds. If you attempted to fall directly down upon a border between face plates your character would be pushed to one side or another instead.