Crowdsourced digital recreation of Olympics 2012?

I managed to get this idea into a tweet of 140 characters. Which I guess means that its not too bad an idea and easy to explain. The original text was
“2012 olympics spectators to share their videos. Pump video into project natal. Mocaped events rerendered in vw. Bypass agencies”

Image from Konami Track and Field at arcade-history.com
To explain the how and the why of this a bit more now the elevator pitch is done.
The London 2012 olympics is a huge event that will be taking place in a world that is already heavily immersed in social media. People are already willing to take photos, share their experiences. A mass event like the Olympics will mean huge numbers of people sharing that experience from their mobile devices at the event.
On the other hand we have an event that has lots of international politics around it, huge sponsorship deals to guarantee coverage and to also pay for the logistics of the event.
We want 2012 to be the best games ever. However we are in a cash strapped world coming out of a major recession.
One of the things I have majored on over the past few years is the recreation of sporting events (and other training events too) from data and descriptions of the event. This is in part a technical interest but also one where I know that sport is understood by most people, so representation of sport in new ways helps them understand things like virtual worlds, social media and games.
I got to thinking that if the various bodies and rights management people controlling the Olympics flow of content and data have the whole thing locked down, negotiated and controlled but are not doing anything forward thinking now using the emerging technologies then we need another option.
A crowdsourced and opensource approach may just be one that at least demonstrates the potential power of the crowd covering the event.
The basic premise.
2012 a high percentage of people able to attend the olympics either on corporate tickets or paid for will have good mobile equipment, cameras etc. Network coverage will be good as it will be boosted for the event too. So there will be a vast amount of tweets, pictures, videos, streaming video, blog posts, facebook chat etc flowing in and around the events. That will of course not be as good as the TV company coverage but it will still have a great deal of potential.
Taking some of those images, videos etc and utilising technology to work out which athlete is where, what event was this, and even motion capture the individuals at events using Project Natal (like) technology means we have a great deal of data available to digitally recreate the games in the public domain. Its the sort of approach Photosynth takes to recreate imagery from mass use of digital media. To feedback to the crowds with data driven augmented reality applications etc. It is a little way out, but it is not impossible and there is a lot of technical and social innovation that can occur in the next 2 years with web, mobile, virtual worlds and games.
I think London 2012 needs a focussed and serious look at the things that can have a huge impact to the coverage and popularity of the games. The things I am considering here are homebrew ideas. They can be done very cheaply by anyone. Their is huge scope for something much greater to be done for very little extra cost and effort in the scheme of things by the powers that be.
Capture, Analyse, Recreate, Share. Yes CARS for the olympics.