20 years ago, the World Wide Web was still in its formative stages. Just a few years later, we had the dot-com boom, as everyone sold off traditional brick and mortar retail names in favor of their internet counterparts, which didn't suffer from nearly as much overhead liability and should've ended up obviating the traditional retail model. Larry Ellison, the CEO of Oracle, captured the spirit of the times in the mid 90s:
"I hate the PC with a passion. Me going down to the store and buying Windows 95, I've got to get into my car drive down to a store buy a cardboard box full of bits you know encoded on a piece of plastic CDROM and you bring it home and read a manual install this thing - you must be kidding you know, put the stuff on the net - it's bits, don't put bits in cardboard, cardboard in trucks, trucks to stores, me go to the store, you know, pick the stuff out, it's insane. OK I love the Internet - I want information you know it flows across the wire."
In theory, this made sense--in the future, everything will likely be delivered digitally. But when in the future? There have been enormous leaps in bandwidth and application delivery since the dot-com bust, but there have also been exponential increases the size of programs as well as associated security complications. My point is, we look upon digital delivery favorably because we understand all of its benefits, but we don't realize the downside of future complications of such a model yet.
Also, there isn't a digital model that can adequately replace the used-game business. I guess the American consumer is just going to change fundamentally going forward, and buy all of his software new and at full cost, without a transferable license. Unless you're one of the Larry Ellisons of the world, you can't realistically believe that wet dream is coming to pass anytime soon.