I have a question - and before I ask, let me be clear that I do not mean to suggest that anything would come of me asking, but... what would you lot consider to be the core features needed in a CM clone? Aside from enough moddability that enterprising souls could replace "generic space empire" with "specific space empire" if they so chose?
Well, back in the old days when Tellurium from the BroQuest team pitched the concept, it was basically "King of Dragon Pass:, 40k edition". It picked up some Dwarf Fortress attributes and other elements even before the first attempt at making it, but for a bare minimum, I'd say that the thing to do is go back to the original concept without getting too into the more 4x type things Duke added.
That means that there are basically three things you need. First and most foundatiobal, you need a management screen, or set of screens, where you interact with important parameters of your chapter. For this segment to be worth it, you at least need ways to organize and replenish your troops and ships to some degree, although managing your fortress worlds and sources of recruitment would also be fun. Duke's version allowed a lot of individual customization of units but I don't think that's strictly necessary. I don't think having a numerical economy like his is strictly necessary for this either. You can have ships be a lot more unique, get them from narrative sources for example.
The second thing you need, which Duke did first if I recall correctly, is a sector. The sector should have systems and planets (which should know which faction controls them) but that's pretty much all it needs. Even having your fleet be at a specific place on the map isn't strictly necessary, you can abstract it to just launching missions from the fortress world.
Which brings me to the third and most important of the main things, and the one in which Duke's game was most lacking - but also the one with the greatest potential for a coder to say "okay, this is the format for this, other people make stuff and post it if you want". That was the core premise for the first attempt at the game, but that one got nowhere because nobody ever coded an engine. Anyway, I'm talking about events and missions. Maybe that's two systems, but that's on a lower level of design. Basically, there needs to be the possibility for an event to happen (presumably on a turn resolve) and for the player to respond to it. In its simplest possible form, the event is "xeno scum attacked world x" and the response is "load infantry company y into ship z and go there and kill them". Now that I think of it, pretty much the same gameplay loop as XCOM.
But that's just at the basic level. Setting aside the high levels of numeric detail that Duke went into, other key ideas (not necessarily the first playable build, but ideas that buids should be written in a way that allows for) are non-combat events of as many types as possible, the ability to have a council of named characters who give you advice (as the nobles in KoDP) and cool shit such as relics, extra tricked out ships, and especially badass characters (who would presumably also be the named characters for the council).
Also, if we're talking about the minimum requisite features.
- able to move ships full of marines from planet to planet
Not necessary at all. Ship location can be implied.
- different planets would be filled with different enemies and battle would commence when a ship attacks it (in a text based log, no need for graphics).
Yeah, this is pretty basic.
- enemies would spread to other planets per turn(s).
Not strictly necessary.
- periodic or action based events with some choice (going to planet Balthazar might give you the choice to enter a tomb and potentially get loot, but you might lose soldiers etc.)
This is the most important gameplay feature since such a system can also define enemy behavior.
- marines could be equipped with different gear which would affect their stats (this could be pretty broad to start with - the equipment could be set per ship rather than individually for instance).
Not strictly necessary. Consider that KoDP has carls, thanes, and that is the entirety of the combat distinctions. Having a few more unit types would be better, sure (and if the pie is in the sky anyway, I'd like Dominions-style tactical battles, but it's not the bare minimum. One unit type and some dice rolling behind the scenes is enough to function. It works for Risk, it can work here. I'd like it if it were possible to define a unit type by specifying exact loadout, and then equip (or re-equip) units to that or any unit type you come up with, but this kind of thing is very much not the minimum requisite.
an inventory for each ship would be needed for instance
It really wouldn't be. The minimum requirement for a ship is the same as a crusader kings ship: It needs to know how many troops it holds. If ships are assumed to be at your base when not in use, then they don't even need to know who's in them. You just sum up the capacity of all the ships being dedicated to the event, and that's the maximum number of marines that you can use.
I'd also suggest it be done in a more global language than Gamemaker - python, C# or similar, as gamemaker can be a bit restrictive and isn't really suitable for githubbing etc.
A language doesn't have to not be terrible in order to be suitable. You could meet the basic requirements in Twine, including moddability. I'm not recommending that, mostly because it's awkward for geographic stuff and gets terribly bloated terribly quickly, but it's sufficient to the minimum requirements.