So today I got Vista's speech recognition software to work. At first I was having big problems, but most of those were because the microphone that I was using sucked. Right now, I'm using a headset microphone, which is a much better. The first microphone I tried was in my little web camera, and it had a very hard time understanding me.
Dictation isn't perfect, even with the headset microphone. But it's a lot better. From what I understand, it will get better as I use it and as I correct its mistakes. I actually think it's much more useful for controlling windows as opposed to dictation. It seems to understand windows commands very well. I'm still trying to develop that is useful, of course. Time will tell.