There has been an interesting public discussion during a recent computer security conference on the real economic value of virtual goods, specifically objects from MMOGs. Here is Game Girl Advance's entry on it.
Reading about this led me to the website of Edward Castronova, probably the best known academic working in this area, and the website of Julian Dibbell, who, after writing an article for Wired on the subject, decided to enter into the business himself, and write about his adventures.
The most significant and groundbreaking aspect of the computer is simulation: the ability that allows one to represent abstract ideas from a process perspective - something which is so hard in other mediums of representation that it has never taken off. In, say, a book, you can either show the output of a process, and even that is quite limited, or a non-functioning description, such as mathematical formulae. Computers allow you to represent abstract ideas as functioning processes, or as a part of functioning processes. Money, which is an abstract idea in an abstract process, is almost entirely represented by computers now. Many more concepts which we have taken to be concrete because of their historical link with concrete things are now moving to representation by computers. And in the process, the historical link with concrete things is broken. As is happening with money, and, in this case, with property.
I'm still too tired to try and make that make sense.