This is the month when controversial xG will, we're told, start to roll out its mysterious xMax technology in Florida.

To keep us all excited, an email arrives with an upbeat analysis from financial conglomerate ING, wrapped up with a separate commentary from Robyn Harte-Bunting, analyst at stockbrokers Hitchens Harrison & Co..

Harte-Bunting sticks his neck out and says "We believe the core technology to be sound while partnership is the key to obviating the execution risk." ING admits it cannot tell if xG's technology can work, but reckons there's a lot to gain if it does - and not much to lose if it doesn't. The company's valued at $1.2 billion, which ING says will go up to $6.2 billion if the company succeeds, and down to $735 million if the technology works but doesn't take off.

But does it work? "To protect its intellectual property, xG has not told us or made public how its Flash Signal actually works," says ING. "Even if we were provided with details, we do not have the technical expertise to undertake a rigorous engineering assessment."

It does say this, though: "xMax is not as spectrally efficient (bits per second/Hertz) as GSM or CDMA, but can send and receive data at lower power levels," and that's the nub of xG's attempts to wiggle out of explicitly making any impossible claims. The second half of that sentence is meaningless - any system can send and receive data at lower power levels. But the data rates go down as the signal power goes down. That's a basic consequence of the the Shannon-Hartley Theorem. There's an implied claim there, that xG can do vastly better than existing technologies at low powers and - since those technologies are doing pretty well efficiency wise - that's a very unlikely claim, simply because of physical laws.

ING excludes the possibility of utter failure, because xG is on the verge of roll-out late in November, but this turns out to be only for "friends and family", with no paying customers till January, according to ING.