I’m very much an idealist so I kind of want to do everything and try make it better. Performance is my main interest, I want to make maximum use of computer power rather than dragging it back with inefficient software.
The problem is while I have a pretty good idea of most of the theory, I have almost very little practical experience, and my idealistic tendencies tend to keep it that way. I’m trying to get past it.
I’m currently trying to write a voxel engine (akin to a game called Block N Load, whose performance issues drove me to this), to see what I can do performance wise. I’ve got a couple of my own ideas to try out.
And there’s no less than ten other projects I want to do, and a few are unique as far as I can tell. I’ll keep those to myself for the moment. 
On to you, even if there’s basically a library for everything, perhaps you can find flaws in them, or maybe there’s something more specific you need, or you could create libraries targetting a more niche audience.
You could also look at web applications. One thing I’d like is a less restrictive project/task management system. Ganterr and the like have annoyed me more than pleased me, though I certainly haven’t seen all projects of the type. I’d also like an online video player that isn’t doesn’t completely break if you click a few buttons too fast, or lose connection for a moment.
There’s also dekstop tools, I’m sure there’s plenty of cool things you do to assist in asset pipelines. You could even make plugins for existing software.
And secondly, is there any application that you want to exist? That you’d find ideal?
