The author hits the drawbacks pretty well. Politicians only care about broad opinion polls because those tell them how to market themselves to actually win elections. Letters and emails, and tweets!, written to representatives suffer from too much selection bias for them to care.
However, this will be a great placebo for the masses of people that like to feel involved. So it could draw a lot of traffic anyway, which could mean money. I can see his site being something that Wolf Blitzer checks constantly the day before the election. Like CNN's moving maps, this can become part of the grand spectacle of Democracy that we have all grown to know and love.
Interesting... the application of twitter and gMaps mashups to politics is something I'm keeping a close eye on for my future races. I especially liked this map that the Corzine people rolled out in NJ: http://www.joncorzine09.com/main.cfm?actionId=globalShowStat...
The follow-up thought that a friend of a friend who started to run for a delegate seat in MD was to design a system where someone could put in a small ~30 sec video clip of why you were supporting a candidate. You could use these map-clip mashups to generate neighborhood meetings, organize precinct meetings, and probably lots more.
If you're in the U.S., checkout 2gov.org - tweet @2gov with your view on current issues and they'll automatically notify all of your elected representatives.
They also provide some pretty cool rollup stats on what folks are tweeting to them.
Really this is something of a response to #1 and #3. Imagine that instead of congress listening, journalists/bloggers/etc used it as a way of doing their own impromptu polls of sentiment on issues by district.
These are separate use cases for the same technology (and same product).
Governmental listening. Media analysis. PR and spin doctor campaign monitoring. Local election candidate passive polling. Local agency (e.g. Scottish NHS) polling. The list is pretty endless.
The problem is that so, so many of these people just don't have big budgets. Good sentiment analysis is hard. Tracking and quantifying opinion across platforms (not just Twitter), in real-time, to an appropriate degree of accuracy doesn't come that cheap - even with web-twenty cloud-malarkey stuff.
We've found other (potential!) sources of funding beyond the government and agencies themselves, 'cos they just don't have flash cash, and with their spending under scrutiny in the UK... yeah, we basically had to.
Taking it further, you could build a vocabulary of issues[...] extract them from Twitter, along with their geo-locations, and do a sentiment-analysis on them [...]
You're making the assumption that Congress actually values listening :-)
I think government transparency is great as an ideal, but what seem to be happening so far is that there's a lot of lip service being paid to show that things are changing, and I realize that it is early days, but Congress "listens" to "polilng" because that is the only way they can stay in power, all other forms of listening have much lower value to them.
However, this will be a great placebo for the masses of people that like to feel involved. So it could draw a lot of traffic anyway, which could mean money. I can see his site being something that Wolf Blitzer checks constantly the day before the election. Like CNN's moving maps, this can become part of the grand spectacle of Democracy that we have all grown to know and love.