Thanks for the heads up!

It’s an ambitious project, but I agree with you that the Google Corpus Data is the best source for analysis.