My first thought is of how many video games there are that rely on text and don't even have voice actors. It also makes me consider procedurally-generated games and how useful a fully dynamic voice with realtime text to speech would be an amazing addition to such a project. Of course, independent and amateur developers could benefit from the tech. Although I imagine there would be a number of them where it would be better to have no voice rather than a poorly implemented AI synth voice.
Such technology would be great for games like Dynasty Warriors with its huge cast of characters, where the English dub of Dynasty Warriors 9 had such a rushed voice recording schedule to save time and money that it turned into a bit of a farce (no fault of the actors themselves).
I have no doubt that synthesized voice acting will turn out no differently than synthesized music vocals. It will carve out a niche for itself, but it won't be taking over the industry. I could see the actors being used for dramatic story events. Then use the synth versions of them for the mundane conversation tree dialog that most people skip through anyway.