Whenever engineering seems too much to tackle, it’s always best to see if you can break the problem down.
What you describe above is by no means simple, as just the different ways all of it could be implemented are complex:
- does the camera rotate with the player on the asteroid?
- Do the player’s controls rotate as they walk around the asteroid?
- Does the camera lock to the rotating asteroid?
- What does the camera do if the player digs down a ways?
- If the player digs a tunnel into the asteroid, does he walk down on the walls of the tunnel?
- Which side of the tunnel does she walk down?
- what if there is a tunnel fork? can he leap to the other side?
- Does the player bottom out and climb up the other side of the well?
- Again, what does the camera do when this happens? etc.etc. etc.
Try to set yourself up for success first: try making a single fixed asteroid in space. Then try out any one of the above ideas (tilemap, bitmap, geometry that you deform, whatever you want to try first) and get ALL of that working properly:
- get the player maneuvering this fixed-in-place asteroid regardless of its shape or size
- decide what the camera might do as the player goes around and under the asteroid
- decide what the controls might do then
- get them actually digging on the part of the asteroid in front of their shovel
- get the asteroid changing in response to that
- etc.
Absolutely NOTHING above will be done “all at once.” Every part of it will be iterative piece by piece, experimentation and refinement.
Do not consider moving onto floating multiple asteroids that you can leap on and off of until you have 100% of all the walking and digging working in a single static asteroid.
Again, set yourself up for success, not complex compounding confusion.
And as Corny said, do not even THINK about optimization. Think about gamejam, gamejam, gamejam, test, test, test, iterate, try, fail, try, learn, try, fail, succeed, layer, the next step on, try, fail, succeed move on,