I think the grants are a great idea, but simply allowing anyone to make a grant address and have it voted on may not be the best way to accomplish a meaningful dent in the crypto world (people will always be greedy, and do anything they can to game the system).
My idea would be to implement a system where grants are proposed (and explained, maybe make a minimum # of words on explanation) through a "grant" tab on the QT miner, and need a certain number of votes (separate from sending satoshis) from mining addresses with >40MEG (or whatever number would work) to become a valid grant proposal, at which point they can be sent satoshis to receive grant status. It still fits the theme of decentralization since there is no central authority deciding what is a valid grant, it would just require tuning of the # of votes and amount of required MEG per valid voting address to make a grant.
If you also allowed down-voting of a grant, it would ultimately give the masses capability to add or subsequently remove a grant if it was decided it should not be receiving grant coins anymore, as I'm sure would be the case if an individual managed to game that system and vote themselves a personal grant like is currently happening.
This system could allow as many grants as were necessary, and grants could receive a weighted # of the total grant money depending on how many votes they have received (this could be limited by only giving each valid voting address a set # of votes which can be used or retracted at any time, or unlimited to allow groups of miners to jockey for small grants). The overall idea is simply to give the grant system some checks and balances, to ensure that people are supporting a grant, not just one person.
If I knew anything beyond basic C++, I'd attempt this myself, however I lack the knowledge or programming experience to even have a chance of success. If anyone can point out any major flaws or necessary refinements in this idea, feel free to. If we can agree on a system that would be fair and have some semblance of balance, I'd be happy to donate whatever I can to anyone willing to undertake the actual assembly of the code.
Beyond that, I'm still having trouble going more than a day or two without the miner hanging on a block...right now I firmly believe that's the primary hindrance to this coin's success. People don't want to have to constantly monitor their miners just to make sure they're still mining, and if they have to, they'll move on to a coin that doesn't require this.