Can Microsoft pull one more rabbit out of its hat?

The story was revealed in sworn testimony from Mark Lucovsky, a former senior engineer at Microsoft. After informing Ballmer that he was leaving for Google, Lucovsky claims, Microsoft's CEO went into a rage, tossing chairs and vowing to "kill" Google.

After watching the big galoot do his monkeyboy dance, one suspects that it's entirely possible that Ballmer temporarily went mental. You so rarely find technology executives of any stature revealing themselves as human beings. Ballmer's different. Microsoft's CEO wears his heart on his considerable sleeve and is not shy about demonstrating his passion--even if that means occasionally losing it.

I'm sure some of that take-no-prisoners mentality factored into Microsoft's decision to block former employee, Kai-Fu Lee, from defecting to Google. Microsoft can't allow this sort of thing to become a habit. Especially not now, with Google replacing Microsoft as the company that most inspires fear and fascination in Silicon Valley.

It's hard to overstate the enormity of Microsoft's predicament. Google is the first challenger in memory with the wherewithal to put major hurt on Microsoft. Last week, Google added another $4.1 billion to its coffers with a secondary offering (on top of the $1.67 billion it raised during its IPO). And considering that it accounts for more than half the search market, Google can bankroll experiments with any number of new services.

Microsoft employed a similar strategy with great success. Throughout the 1980s and 1990s, the company's Windows and Office businesses were veritable money machines. Microsoft could keep hacking away on other projects until they were ready for prime time because it enjoyed a lucrative operating-system monopoly.

Along the way, there were pretenders to the throne, like Novell and Netscape--and we all know how they ended up. But Google won't depart so quietly into the night. When Ballmer looks in the mirror, he sees Eric Schmidt's reflection. Google has become the living embodiment of the much-ballyhooed concept of software as a service, with the constant flow of new products onto its Web page as beta projects.

What's more, this isn't the old Microsoft that Silicon Valley once dreaded. Behind the company's sweeping reorganization this week is the recognition that the current system is in need of a major lube job. Above all, products must get to market faster.

When I toured the company's labs last year, one Microsoft engineer bemoaned how a pet photo display project wouldn't be implemented until the next major version of Windows was ready. His frustration was compounded by the knowledge that Apple Computer already had a similar offering on the market.

Google's chances also depend on the PC losing its primacy. Oracle's Larry Ellison argued as much when he was out pushing his idea of a network computer. So did Sun Microsystems' Scott McNealy with his mantra that "the network is the computer."

Ellison and McNealy made little headway. But they were just too early. The Internet has scrambled old assumptions by becoming the new platform. How long before everything resides in the "Internet cloud" rather than terrestrial PCs and servers? Maybe not this year or next--but five years hence isn't a stretch.

At the same time, there are now lots of alternative ways to access the Web besides a Windows-based PC. That spells trouble for Microsoft's monopoly franchise, which depends on maintaining the status quo.

Try imagining a future where developers will write to Web platforms without thinking about an individual computer or operating system. That once was Netscape's dream. If this does come to pass, Google could build an ecosystem around itself in much the same way Microsoft did with Windows. If Microsoft's latest moves can't clear out its corporate arteries, the future could be all Google, all the time.

And at that point, you'll really see Ballmer pitch a monkey dance fit.