It turns out those envisioning the internet as having atransformative effect in the 1990s were not crazy dreamers. Rather, they just underestimated how long it takes for technology to have impact. Society does not move on "internet time."
PEW issued a report documenting the trend. The number of people using the internet as their primary source of political news in 2006 nearly doubled from the previous mid-term election in 2002, from approximately 8% to approximately 15%. http://www.pewinternet.org/PPF/r/199/source/rss/report_display.asp
That's still fairly low as a percentage of the population overall, and it does not diminish the critical importance of local broadcast and local newspaper coverage. Indeed, it is interesting to note that while reliance on newspapers declined (in synch with the overall decline in newspaper readership), broadcast television remained dominant as the means by which people got their news.
(Also of note is that more people rely on broadcast and newspaper outlets in a Presidential election year than in an off election year. This suggests that user preference may be driven by the utter dearth of local election coverage.)
Nor has the internet yet eliminated the power of local party machinery and the other numerous elements that give us our large rate of incumbency, low voter turnout, and other issues that bedevil us as a functioning democracy. But even here, the trend line is encouraging. Of those using the internet for political news, 23% also used the internet for more active means of civic engagement.
As I have said with regard to numerous other trends: "you will not even notice it until it is all around you, then it will have been inevitable." There is nothing that says the internet must inevitably transform politics and civic engagement from the passive and empty rituals they became in the late 20th Century into something more akin to the notion of the public sphere envisioned by Habermaus and others. It's one of the reasons I care so much about network neutrality and other issues that impact the openness and ubiquity of high-speed networks. But, by the same token, it is not inevitable that these technologies will fail in their promise.
It is entirely up to us.