Voice recognition has always been patchy or non-existent for languages other than English, and it's a deal breaker for something that relies _primarily_ on voice for input. That's the point I believe he was making.

I don't think he was referring to the glasses in specific rather than the state of voice recognition as a whole.