When Google (Nasdaq: GOOG) launches a high-flying business experiment, it can be a magnificent success -- like the original, simple online search engine that made the company, the auction-style advertising model that made it rich, and the up-and-coming online office suite that makes Microsoft (Nasdaq: MSFT) sleep with one eye open.

There's genius everywhere at Google, but until they turn pro, the ideas are like popcorn in the pan. Some pop ...

... some don't
The grand experiment in telecom sales known as the Nexus One is going down in flames. Google launched a website this January to push Nexus One phones out to consumers, making Google the point of contact and launch partner T-Mobile from Deutsche Telekom (NYSE: DT) an afterthought.

Premium program partners were lined up out the door. Motorola (NYSE: MOT) would provide future Google-branded Android phones; Sprint Nextel (NYSE: S) and Verizon (NYSE: VZ) Wireless said early on that the Nexus One would soon be supported on their networks. This experiment looked sd if it might pop.

Houston, we have a problem!
It didn't take long to find cracks in Google's armor. Early adopters soon came asking for better tech support -- emails and Web forums were never quite enough. And many a potential customer never pulled the trigger on that purchase, because those customers had no way of knowing what the phone would look like, feel like, or do for them.

Meanwhile, the Droid Incredible from Verizon and HTC already made the Nexus One look obsolete to the Verizon faithful, and the HTC EVO 4G did the same for Sprint fans. Unsurprisingly, then, Verizon backed out of its Nexus One commitment -- followed by Sprint. And now Google is getting out of the direct sales game altogether, planning instead to ship Nexus phones through traditional distribution channels like stores in the neighborhood mall. That not-so-old online site will lose the "buy" buttons and turn into a mere showcase for Android phones in general.

The good and the bad
So the Nexus One is back to what I always imagined it to be: just one of a gazillion very capable Android handsets that just happens to bear a Google logo. The company still plans to continue its hardware project, possibly with something like a Nexus Two showing off what newer versions of the Android software can do, but only by bog-standard sales vehicles like service provider partnerships.

As an American consumer, I'm sad to see Google failing this attempt to work around the middleman in the phone-buying process. But as a Google investor, I don't really care much. The Nexus One was never a huge contributor to the top or bottom lines of Google's business, and it wouldn't hurt much even if the company would stop making phones altogether.

The name of the game is to improve mobile technology, which increases mobile Web use, which in turn benefits Google through indirect means. Whether you click a Google ad on a Nexus One bought online, a Nexus you bought from your local T-Mobile store, or an Apple (Nasdaq: AAPL) iPhone is immaterial. Google makes money either way.

What have we learned here?
So the Nexus One kernel never popped, and its innovative distribution model was the wrong tool for this job. On the other hand, the bigger Android kernel is turning out to be juicy, and smartphones look destined to become the new "standard" phone in a few short years.

If the failed online-only experiment in expensive personal electronics has taught us anything, it would be that the service providers hold a lot of the cards in the phone-buying game. Google's ambition risked turning their model upon its head. Choose a phone first, and then make the provider a secondary thought. However, as Google learned, changing consumer preferences is a difficult trick, and well outside the Big G's area of expertise. If anything, the whole experience shows the power Apple's fancy retail stores may have been very key to building the iPhone and smartphone phenomena.

Feel free to discuss Google's failings and victories in the comments below.