Diversification Doesn't Work Anymore?

I've seen a number of articles recently that bemoan the fact that diversification among varioius asset classes doesn't seem to work anymore. Typical of these articles is this one by Jim Juback at MSN Money:

I thought investing in gold, copper, zinc and other commodity stocks was supposed to diversify a portfolio and protect it from falling too far, too fast in a general stock market retreat.
Some protection. Maybe instead of calling it "diversification," they should rename it "di-worsification." Portfolio diversification, that tried-and-true tool for reducing the volatility and risk of a portfolio by making sure that some of what you own will go up even if the rest of it goes down, isn't working too well right now.

This seems to me a vast misunderstanding of the principles of diversification. I don't expect that diversification will protect our portfolios from every downturn in the market. Indeed, during corrections driven by changes in liquidity, it seems apparent that most asset classes will move together. Furthermore, Juback seems to have forgotten the one asset class that did do well during the selloff: bonds.

All of our portfolios except the most aggressive contain bonds and/or cash. Diversification worked exactly as it is supposed to during this correction. The bond portion of the portfolio gained while the risk assets declined. The percentage of the portfolio in bonds determined the degree to which your portfolio was protected. What's so unusual about that? Nothing. Juback seems to be upset that a portfolio with all risk assets declined when investors wanted to reduce risk. Duh.

Diversification works just fine when applied properly. It also tends to work better when viewed from a longer time perspective. Even a portfolio of all risky assets will do fine over a longer period of time. Expecting a diversification strategy to also protect you over a short time frame is unrealistic.