There is really no way to convey the sensation that awaits you by my typing mere words here on your screen.
Go NOW to Yamaha’s website at http://www.vocaloid.com/vocaloid4/lineup/cyberdiva.html to experience the future for yourself! Listen to the demos! Can you even tell it’s not a real person singing?
And DO NOT miss out on the Demo here:
I don’t know how this compares to other vocaloid technology, but I can hear it’s not a real person within a few seconds…
It’s not my cup of tea, I would never consider this as an alternative to a real vocal.
That said, I’m sure it’s a capable creative tool and fun to work with, so by all means I hope you enjoy using it 
The demo in the second link Sachiko was much better than the diva ones… but still too far from real. Cool though 
There is this this term most often used in 3D modeling, humanoid robots and related fields, “uncanny valley”, where the valley is a dip in a curve on a graph over how similar robots and humanoids look the real thing. the uncanny part is how much the model creeps humans out. It’s an interesting thing you can google if you want to.
What do I think of Vocaloid? Well, I’m kinda clueless, that’s all. 
Well, you guys are all nicer than I am!
I was being sarcastic in the original post. When I first clicked on the demos, I just started laughing pretty hard. I think Siri is probably a better singer!
Anyway, you guys are so nice about it, I kinda feel bad now… 
Ah well, I thought you were serious and didn’t want to rain on your parade
I stand by what I said though, I’m sure it’s a capable tool and you can do fun stuff with it, I just wouldn’t consider it a replacement for a real vocal.
And I agree with NWMW, Sachiko sounds a whole lot more convincing than cyberdiva.
My inner “physical modelling geek” loves this stuff, Synths may still rule the planet yet 
Beware of the Ulfaloid 
The uncanny ulfaloid, that sounds like it would sell well!
Yep! Much better than the uncanned strophaloid!
Although most of the demos I heard sound cheesy With the exception of a few demos that sounded good/human. It might work if you can get it to sound more human. Like this demo:
I’m not sure but the song: Don’t be gone too long by Chris Brown has a female voice singing in unison that sounds like it could be a good fit for it.
Overall I could see it work for duets, background vocals and vocoder style vocals for electronic music with a real vocalist.