The right way to think about these figures is in Schumpeterian terms: With retail sales down across the board, whose businesses are being destroyed here, and what is the future of physical retail? Amazon is merely goosing this process along, of course, and may not even end up being a survivor.
Such is the nature of business: some people lose while others win. It's not exactly a zero-sum game, but it can sure feel that way at times.
While the Web lobotomizes traditional retail, it's doing the same to software: open source, SaaS, and Web 2.0. I've been complaining lately about Web 2.0's effects on the media industry, but my concerns are probably overblown...in the long run.
The common denominator in all this creative destruction is the Web. It's not source code, data, or really anything else. When you analyze open source, Web 2.0, Amazon, etc., the real game-changer is the ease of access and distribution via the Web. Everything else is largely frosting.
What's the one big thing that Microsoft struggles to replicate? Not data lock-in. It has had that in spades with Microsoft Office file formats. Ditto with the positive (for Microsoft) network effects deriving from everyone using its tools/applications. Heck, Microsoft even arguably has some of the Web 2.0 benefits of self-improving applications as more and more people use its software, reporting crashes, bugs, etc.
But what Microsoft and the 1.0 world don't have is free distribution, free access. That is the thing that is roiling old-world businesses and replacing them with new-world businesses, ones that, for the most part, still haven't figured out how to make much money--with exceptions such as Google, Amazon, eBay, etc.
But if you take a close look at these companies, you notice that while they absolutely do display some of the hallmarks of the 2.0 world, their money comes in distinctively 1.0 ways: selling goods for lower cost, higher volumes, and at greater efficiencies--efficiencies enabled by effective use of the Web for cheap access and distribution.
Going forward, I think we're going to see more of the same: open-source companies using the Web to efficiently seed the market and distribute software, but with largely traditional 1.0 value on the other end of the phone. SaaS providers doing the same, but providing even greater 1.0 lock-in via centralized distribution over the Web. Web services companies using the Internet to aggregate and orchestrate content but ultimately paying or otherwise centralizing the best authors (Wikipedia, anyone?).
The Web has changed the economics of business, just not as completely as we had assumed. It has a dramatic effect on the price of distribution and customer discovery. As for actually delivering the goods, well, that's still mostly a 1.0 affair, which is a testament to the model's power. It can borrow from the Web while still monetizing off the Web.