By Paul Keegan, contributor

It was an awesome spectacle as product launches go: Speeches by Arnold Schwarzenegger and Colin Powell, a panel of top executives from Google, eBay, Wal-Mart, FedEx, Coca-Cola, and Cox Enterprises, video messages from Diane Feinstein and Michael Bloomberg, a heart-tugging slide show about saving the planet, and finally, after eight years and $400 million raised, the unveiling of the Holy Grail in a box — clean, inexpensive full-cell energy!

“I would like to introduce to you the Bloom Energy Server,” said Bloom CEO K.R. Sridhar to applause yesterday as colored lights bounced off the hulking black box on stage. “This is my baby, isn’t she beautiful?”

So did they pull it off? Does the Bloom box live up to its hype? Let’s let board member Colin Powell answer. He promised his wife he would install one of the energy-saving devices in their home instead of a conventional generator, he said, adding, “She’s still waiting.”

So are we.

The cornerstone of the presentation at eBay’s (EBAY) corporate headquarters in San Jose, California was supposed to be Bloom’s happy corporate customers — 20 Fortune 100 companies in all — but a quick survey of some of them reveals that, in many important respects, the Bloom box remains in stealth mode.

Coca-Cola (KO) says its first Bloom fuel cells probably won’t be installed until September. Bank of America (BAC)? Sometime later this year. FedEx is close but still not online. Two firms who actually have experience with the Bloom Box since the first units were shipped 18 months ago — Google (GOOG) and Wal-Mart (WMT) — declined Fortune’s interview requests.

FedEx (FDX) did agree to talk but won’t answer the most important question: How much are they paying for Bloom-box electricity? “We’re under a contractual agreement not to disclose the price,” says Mitch Jackson, FedEx’s senior director of environmental affairs.

Okay, so let’s do the math ourselves: Bloom says each 100-kilowatt Bloom box is sold for $700,000 to $800,000. Factoring in the price of the natural gas that most often fuels the unit brings the cost to 8 to 10 cents per kilowatt hour. Hey, that’s cheaper than the 11-cent average retail cost of electricity you can get from the coal-powered grid!

FedEx ran the numbers, too. The five 100 kilowatt systems it’s installing at its west coast sorting hub in Oakland, at $750,000 each, would cost $3.75 million. But the energy production is so cost-effective, Jackson says, “It should pay for itself in less than five years.”

But look at the fine print: That 8-to-10-cents per kilowatt hour includes huge government subsidies — a 30% federal tax credit and $2,500-per-kilowatt rebate from the state of California (see editor’s note below). Take out those subsidies and the true cost of the Bloom box is 13 cents to 14 cents per kilowatt hour, according to analysis by Lux Research — significantly higher than electricity from the fossil-fuel powered grid.

Grid power also doesn’t require the headache of owning your own power plant. Yes, Bloom offers a 10-year warranty and will maintain the unit, but if the system fails — and reliability is a major issue for fuel cells — companies would have to plug back into the grid, which for all its drawbacks remains enormously reliable. A replacement for the grid? So far, it’s more like a supplement at best.

FedEx offers a compelling example: The company says its new Bloom boxes will be enough to provide about 55% of its facility’s energy needs. Another 20% comes from 5,700 solar panels. The rest will come from the grid, which also serves as a convenient backup should the Bloom boxes fail (performance guarantees are built into the company’s contract with Bloom, Jackson says).

So when can FedEx start installing Bloom boxes in the vast majority of states that don’t offer subsidies? “That’s a question for Bloom, not us,” says Jackson. “Any of these technologies, including solar and wind, have to be cost-effective on their own to be commercially viable over the long term.” In the meantime, until the economics work out, at least FedEx employees can help the environment: The Bloom boxes will cut carbon emissions at the Oakland facility by 30 to 35%. “If you combine Bloom boxes with our solar electric array, we should reduce our carbon footprint by around half,” Jackson says.

So if Bloom’s magic boxes are still not cost-effective nationwide — if they remain a niche product still in beta testing — why today’s star-studded coming-out party? Sridhar says his customers were so excited by the product they pushed him to unveil his creation now. Though they were clearly jazzed about the chance to brag about their green cred — and moments at yesterday’s event devolved into an orgy of self-congratulation — many observers think Bloom had another motive: To prime the pump to go public.

“From Bloom’s point of view, an IPO is the number one priority,” says Jacob Grose, senior analyst with Lux Research, who says the success of last year’s public offering by lithium-ion battery maker A123 showed that the market may be ripe for another clean-tech play. “I’d be shocked if this were not a prelude to an IPO over the next 12 to 18 months.”

So after eight years and $400 million, Sridhar needs more time and more money — perhaps another five years to make the Bloom boxes cost-effective, analysts say. For the planet’s sake, let’s hope he pulls it off. Until then — after all the hoopla — the bottom line remains the same as it was last week: The Bloom box remains a promising idea whose time is yet to come.

An earlier version of this story incorrectly stated that California’s rebate is $2,500 per kilowatt-hour. The correct figure is $2,500 per kilowatt.