Basically saying "I'll work on it until I don't feel like it anymore" doesn't help either.
Unfortunately that's likely to be true of anyone picking this up. Anyone with the skillset to approach this in a realistic way has more profitable ways to spend their time; heck, I optimistically have about half the skills I'd need to do this well and I'm still too busy with software I'm actually paid to write and can get published with and so forth.
Even if you got enough infrastructure together to enable the kind of incremental improvement that amateurs can do in groups, there's still the 40k issues. It's legally radioactive, but more than that, the 40k fandom is just not a fun one for whom to ponder developing software. Every creative decision you make -- every statline, every graphic, every name -- will be ripped apart and viciously mocked by some contingent of grognards for whom it represents the epitome of Where 40K Went Wrong, at which point they will be volubly opposed by grognards of an opposing opinion and every request for feedback will be overwhelmed with skub.
So, yeah, everyone's going to work on this until they get tired of the greentext on /tg/ calling them an idiot and then they're going to drop it. If we ever want to see something like CM, we'll need a framework that can withstand people constantly getting burned out and hating the project and everyone involved with it.