You write: If kept in a vacuum insulated tank (big Thermos bottle), it could last a fairly long time. As the compressed (even liquified at that temp?) air heated slowly it would have to be bled off.
uh, you are joking right? minus 100 degrees centrigrade? can you even comprehend how much energy would be needed AND insulation to keep it at that temperature? Mind you this air is compressed to 4,500 pounds per square inch! So you want to bleed it off as it warms up? Then you might as well keep it hooked up 24/7 to an air compressor that will compress to 4,500 pounds per square inch!!! That is more than the preassure at the bottom of the deepest part of the OCEAN!
It wouldn't need "a lot" of insulation any more than a Thermos bottle or Dewar flask does. The vacuum space between inner and outer vessels can be very small.
The article referencing using air at -100C is five years old. I may have missed it, but the most recent article stated that the air tanks are made of thermoplastic with carbon fiber wrapping. This is very similar to current technology in firefighter SCBAs, except I think the interior liner for firefighting air cylinders is made of aluminum. Also, on their page, they note that the cylinders are designed for use at temperatures of -40C. There's also no mention of using liquified air, and since both oxygen and nitrogen liquify at around -130F, depending on pressure, it looks like they've gone to a straight compressed air system at normal temperatures.
One thing that puzzles me is that according to the article, the tanks hold 100 liters of compressed air at 4500 psi. That only translates to 3.5 cubic feet. A one half hour carbon fiber cylinder for firefighting holds around 45 cubic feet, and weighs less than fifteen pounds. Possibly they mean the tanks have a liquid volume of 100 liters, which would probably translate to being around ten times the size of a 1/2 hour firefighting cylinder.