Paul Tansley: In a perfect world, this new idea would work. It would be possible to check some central database of images - and cross check the ownership of every image ever published online. However, we don't live in a perfect world. The technology just isn't there yet.
I'm a professional photographer. I post an image on Facebook - I already know that there is a good chance that FB might share it with others. I watermark it and embed metadata into the image - starting my copyright. FB respects that copyright metadata on import. However, when someone shares that image, it strips it out - my first level of protection is lost. Next, someone crops out or retouches my watermark. Now the image has no identifier on it at all. So, when a business comes along and wants to use that image for their latest advertising campaign, some government body will try and track down the owner of the image - but they have no way of doing that anymore. It's been stripped bare.
That's the problem.