Waaaaay more specific details and less open-ended or vague bits. A proposal isn't a sort of wish list it is like a requirements document for a project, much like you'd encounter later if you start a career in IT. As such it needs to be very specific about what you aim to do when, how, why, with what tools / existing stuff, etc
Thanks in part to the open ended nature of the proposal it is likely to grow to several GSOC's worth of actual effort. This will become more obvious the more details you add. I would suggest focusing on three large items then cutting them into pieces and moving everything else into stretch goals. For instance: Boats, Minimap, and Journal. Those three could each easily take a month of effort.
Tons and tons of feedback has been added to the proposal since last, including some by me right now.
Overall the proposal is vastly more realistic now, good job. There are still some assorted quality issues in the various descriptions, but we get the idea of where you're trying to go. One highlight is that boats is likely going to be more involved than the compass or minimap, therefore it is likely to take more than 1/3rd the timeline, which is fine. You could do something like 3 weeks - 3 weeks - 6 weeks
While you can still improve the proposal a fair amount you might want to try spending some time at this point sorting out and fixing some existing (related or not) bugs in JoshariasSurvival and/or pull in the Nautical module and try to make it more functional before the app deadline. Seeing you able to address issues in related code is a huge factor in acceptance