It's a great idea, I tried it in German (fairly fluent) and it seemed pretty legit. But I'm guessing you're at the 90% completion point and that last 10% will require another 200% effort from you guys...

What do you use for speech synthesis? I have a little side project I'm investigating text-to-speech engines for (aiming for as much realism as possible, at least in English, plus the ability to possibly modify errant pronunciation somehow, and add pauses of varying length)