Reviewed, and RatMoleRat added in a bunch of comments too
I don't want to discourage or make you feel bad here but your proposal has very fundamental problems with it. It still reads far more like a suggestion thread than a GSOC proposal. You need to identify exact things you aim to do and how you aim to do them. It needs to be almost entirely code with extra content/artwork being little side items not primary goals.
You need to be looking at a lot of the related code to get better ideas on exact implementation goals you'd need to achieve. That might be tricky to do this late in the process but GSOC or not we (and the Oreons) appreciate your attention in the area and hope to see more of it