Woah, there's a lot to dissect in this thread!

Personally, and I think I've said this before (and others may have too), while having more granular categories may, in theory, allow more people to win awards, we're working with a relatively limited voting pool, so I don't feel that having more categories will result in more votes. And some of those who
do take the time to vote may not be inclined to weigh the merits of nominees in "lesser" categories in a fair and complete way. Simply put, whatever they've decided as their favorite game will automatically be nominated in multiple categories, regardless. We're also working on the assumption that everybody who votes in the awards is equipped to judge "excellence" in writing, or music, or design, when all we can likely be sure of is which games they enjoyed playing the most, or perhaps what music the thought was pleasant to listen to.
There's always been (friendly) conflict about what the awards actually
are; a "serious" attempt to highlight the creative "best" of the years games and developers, or a fun popularity contest. Some of you may ask why it can't be both? Frankly, I'd say it's never truly achieved either, instead falling into an awkward middle-ground that still results in yearly discussions on how to "improve" the system.
If we did rewrite the rules to differentiate between, for instance, original characters and preexisting characters (or public-domain characters), we would probably need to provide the extra manpower required in deciding whether certain games meet certain requirements to be eligible in certain categories. I don't want to speak for the people currently invested in producing and maintaining the awards, nor do I want to rule out anything, but I don't think we have the numbers to support expanding the awards to cover such changes.
The truth is, we have relatively few people willing to put in the time to produce something off their own backs for the good of the community (and many thanks to those that do), so caveats are probably going to be unavoidable.
I know this all sounds rather dark, so for the sake of adding something productive, here is what I might do in regards to non-original content (in this case music) being used in games:
A game is eligible for a category if the material in question was created specifically for the game, or, if not created specifically, has not appeared before in any form, and is the credited authors own work.So, a game with music by Kevin McLeod, using tracks freely available on his website, is
not eligible for "Best Music". A game containing unreleased tracks previously composed by Kevin McLeod, but given or licensed
exclusively for use in said game, or tracks
specifically composed by Kevin McLeod for use in said game,
is eligible. If said game wins the award, the award will be credited to Kevin McLeod, and not the developer of the game.
...
I mean, it seems logical to me, but we're never going to satisfy everybody. We could
try to cover all angles, doubling or tripling categories to cover original/preexisting/remixed works, but...that seems kind of redundant? We'd be splitting the numbers of games per category even further than we are now, with single nominees in certain categories.
Personally, and probably realistically, it would make sense to exclude games that use freely available content from most of the technical categories, though they would still be eligible for others (Gameplay, Writing, etc) if these aspects are original.
For the sake of arguments, I'd remove "Best Character" altogether.