That is what a render farm is. A render farm uses the cpu and gpu of each computer in the farm to distribute the load, allowing for more rendering threads to be processed at once. An effective farm can take a project that would take a single computer 24 hours, and can cut that down to 4-5 hours depending on the processing power of each computer in the farm.
Edit: I just noticed you said supporters collective cpu's. That would be extremely expensive to do, as the bandwidth consumption would be insane. A render farm is normally multiple computers in one building (usually all in the same room) on the same LAN network so that you aren't having to send insane amounts of data through the internet. Also, the more varied the processing power of the computers in the farm, the worse of a farm it is, as the host computer has to divide up the threads to send to each computer, and multiple processing speeds would increase the host load. Too much variation in processing power can actually make the rendering process take longer. Most render farms are purpose built, with each computer built with identical specs to minimize this load.