“There should be a more sophisticated damage model” – I concur. We’ve got complex ship model (12 variables!) but the damage model doesn’t match. Perhaps instead of a single “condition” variable we should have 2 or 3: hull condition would be what the current condition is nowadays, and the others could keep track of, say, state of propulsion (engines, reactor, emitters) or structure (bridge, crew quarters, cargo bay etc.).
Then we could do two potentially interesting things. First, we could implement repairs to the individual parts/systems as separate actions, e.g. “repair the engines”, “repair crew quarters”. I’m against replacing the whole part “every other or third repair” though - think about the hassle of keeping track on which repair you are. Instead we could just require some of the constituent materials, e.g. if you have an LCB, you have to replace a bit of AST and MFK every time. I like this better, because (1) it’s simpler to balance, (2) creates public demand for stuff that’s often hidden from the market due to vertical integration (7d units of AST across all CXs: 500).
Bonus feature 1: upgrading the ship to a different hull material could now repair the hull, just like one would expect 
- but not other things obviously.
Bonus feature 2: this way we don’t kill new players. Since “repair the engines” is something you do much less often than “repair the hull”, this could be balanced to only become a problem when the new player has already learned most of the game and got some money.
Second, we could implement more interesting damage mechanics, beyond “micrometeorites hit stuff and everything breaks down linearly”. How about radiation impacting crew quarters and bridge but not cargo bay and hull, while engine damage only scaled with burn time and intensity? How about reentry heat working differently than micrometeorites (maybe different plates could suit different use cases)? How about shields actually shielding selected parts from damage?
