http://www.youtube.com/watch?v=OmUM-MlA8Y4
Haha. I tried to find a good voice synthesiser video, but I don't really know what exactly to look for. But everything I've heard so far in this area, has been nowhere near the quality of speech a human delivers. Yes, they can help a lot when it comes to introductory videos at companies, some speeches for coaching (though the professional ones would still use a persuasive human, because speech-manners also matter there), train announcements etc.
But can they truly convey emotions and be characteristic for a specific character? Include delicate nuances as a natural sigh here or there, an unexpected squeal, some speech mannerism that either the actor has or has made up for the given character?
I don't know, you tell me, as I said, I'm not very oriented in the contemporary speech synthetizers, but I can honestly say, that I doubt it.
As for Hatsune Miku. I played an YT video from some concert to see what it's about, and while I'm in awe for the technology used to make it, I can still hear that there is something lacking. Maybe an emotion, maybe personal touch, maybe an actual depth in voice and a more natural flow (even though, like I said, it amazingly well done). Still, I would always pick a real singer over that one. But I wonder how would she sound singing songs that I actually like

Cause the music I found with her was unbearable for me. And hence could skew my judgement a bit.