Right now I am still doing some interface stuff to send to Nik, but thinking about rescheduling the rest of the semester. I just talked to Mark Chang, our advisor for the project, and he said that we should consider breaking down everything into much smaller tasks, so that people have an accurate idea of how long it will take and we can plan accordingly. He mentioned that doing this has the benefit of not allowing you to say "yes, I'm done" when you have some large chunk of an ill-defined deliverable done.
In a way, it's sort of like test-driven development for deliverables. Before you even begin, name each png you will submit to the art directory, write down the names, arguments, and return values of each function in your library. Block out everything as soon as possible.
But that problem with that issue is a time-based one. We would rather have a single cell with three organelles in a sparse environment working. In order to do that, we don't need to spend time planning out all the organelles, all the environments, and all the other details. However, there's no guarantee that planning out all of those things wouldn't mean some changes need to be made to the first skeleton that we do want to implement, and thus the time should be taken to go over everything at the beginning.
So there are difficulties; this much will be certain. But I feel right now that my vision for how this game is going to progress to completeness is lacking. If I felt instead that I had a good vision but we were irrepairably behind schedule, that might mean we needed to cut out features and go straight to the minimum deliverable. But since vision is lacking, I'm going to use this time, not worry about what deliverables I produce, and instead try and figure out how the rest of the semester is going to be used.