I suspect if you mined Twitter, you would see a 'dumbing down' of English due to the constraints of Twitter. How one would quantify 'dumbing down' is a challenge (if it hasn't been solved already). However, I am more interested in how this 'Twitter effect' cascades to other areas of English.
There are other implications of the internet too I'd love to study but these questions are relevant to this article.