Solar Balloon Progress?

In July 2006 I sketched a solar-energy balloon notion: build a balloon which is roughly half transparent and half (inside) reflective, put a solar collector (I was thinking of solar Stirling engines) around the focal point, give it a supply of water to split into hydrogen, let it fly. I thought of this as part of our semi-Singularitarian notion of the Spike, by which human labor costs become a hyperbolically diminished factor in a world of self-reproducing machines. But maybe it's easier than that, and will come sooner. EcoGeek reports on "Inflatable Solar Arrays" being developed now:

The device uses very little actual photovoltaic material, and everything else is dirt cheap. The installation of something like this could even be several hundred feet off the ground, if one didn't want to disturb habitat.

How about several miles off the ground, to be above the weather? They'll last longer, and be more reliable as power sources. Is flight control really that hard? I don't think so: consider YouTube - solar airship:

This vehicle runs on sunlight, and thus can fly as long as the sun shines (which, of course, it always does above the clouds). You can see the panels inside (see hyperblimp.com for more info).

That assumes a human with a remote control, but we do have people working on solar-powered airships to stay up for "a week or more", as in Danger Room's Commandos Get Hydrogen-Powered Drone

Global Observer, powered by hydrogen fuel cells, will fly at 55,000 to 65,000 feet, serving as a communications' relay for commandos -- and keeping watching over the ground below.

That's a plan, but already we have Slashdot's report of Solar Craft Flies Through Two Nights:

A solar-powered, unmanned craft has flown for 54 hours,

and open-source UAV flight-control is coming right along at Chris Anderson's Do-it-yourself drones. I'm feeling optimistic. This morning they announce the winner(s) of the robot car "race":

The three other cars that were out on the DARPA Urban Challenge course when the three front runners came in have now crossed the finish line. That makes 6 cars that completed the 60-mile course, out of 11 starters.

Their task was immensely harder than getting a solar-powered blimp to lift off, ride the upper winds for a day, and land -- but even that's not needed, if we leave the power stations to fly gradually around and around the world, with manned airships to tend them, supply more water, collect more hydrogen...(or whatever fuel turns out to be the best way to transport the solar energy.)

Faster, please.