I spent a few years of my life working on eMarkets. It was a business that created a ton of hype, most of which came to naught as the protagonists were exceedingly ahead of their time, and even (gasp!) poorly thought out and even more poorly executed. I've always admired eBay for building a phenomenal market out of of exactly the kinds of customers that your MBA marketing classes would tell you never to go after -- the flea-market, short-a-dollar, nickel-squeezing types that make up the collectors, traders, antique dealers, and general sharp bargainin' eye-for-a-deal secondary market of the world.
eBay has done this so well, says the UK Guardian (new bastion of cutting edge tech analysis?) that the average wealth of a UK household has increased by 3000 pounds (about $5000 at current dismal exchange rates).
I suspect that increase in value is almost entirely attributable to two factors; a) the ability to monetize previously unsellable "long tail" items and b) the ability to maximize one's selling price for many "big tail" items that previously might have been sold at a discount for want of multiple competing buyers. Both of these are different aspects of a decrease in transaction costs, most particuarly the costs of locating a willing buyer, or (the flip side of the same deal) the buyer's costs of locating an appropriate item to buy.
This is probably a good time to go sideways into policy and plug Hernado de Soto's fantastic book The Mystery of Capital, which analyzes how a lack of clear, efficient system for allocating property rights in many countries prevents people from borrowing against the value of their homes, and causes cradle death for capitalism.
It's kind of startling to stop and think that the ability to borrow against one's home is a fundamental basis for making our capitalistic world go 'round - and that liquidity is reflected in reasonably efficient mechanisms for turning a house into cash, or vice versa. eBay has taken a whole second tier of capital goods -- mostly what government statistics bureaus are fond of calling "durable goods" but plenty of other stuff, too -- and creating a market of sufficient liquidity that rather than being a straight purchase, the acquisition of many items can now be thought of more as a financial transaction where you go "long" the item in question against an expected curve of depreciation vs. usefulness, with an exit at some future date for a signficant fraction of the item's current value.
It's these kind of tertiary unexpected effects of the Internet that make world-changing techno-social systems so cool.