I think that game ideas posted by you & Guardian are very original and would make a great story-base for a game but... Pappenheimer has a point about what is expected to be presented here.

I guess that, when considering what project would be best to go with, people who will be evaluating these projects need to see several very important aspects like: Will the game be 2D or 3D oriented and to what extent? How many levels will the game have? How many models will be sufficient to populate the game-world? How will game battle-system look like (if it has battles)? Will the game be single-player or will it have hot-seat or multi-player option? Should levels have segmented or non-segmented loading system? ...and many other.

Only thing that I find not to be possible to predict is Milestones, since no one knows how big the game-team will be and what skills will be the team's weaknesses?