I would like to suggest implementing Text to Speech synthesis. This could be used to test dialogs, pieces of information and other that the game intends to have voice actors for. This could help to try out for example dialogs to get a better feeling for how it would sound. This way you can better catch bad dialog earlier that may look good in text, but doesn't work that well when spoken.
There is an online demo here of what I mean:
http://www2.research.att.com/~ttsweb/tts/demo.php
More on MSDN here:
http://msdn.microsoft.com/en-us/library/ms720410(v=vs.85).aspx
and also an open source version here (not as good as Microsoft's version. X11 License, so it should be OK for you to use (No GPL ), but have a look on the license anyway ):
http://festvox.org/festival