If for no other reason, consider the amount of energy that would be released if a tank with enough compressed air to propel an automobile for 50 miles were released in a collision over the course of a few hundred milliseconds.
Say the energy output is equivalent to running a 50-horsepower engine for an hour. 50 HP-hours = about 35000 watt-hours, or 250 megawatt-seconds (MJ).
A stick of dynamite produces about 2 MJ. So that tank of compressed air, if it ever ruptures, will yield the equivalent explosive power of more than 100 sticks of dynamite.
Gasoline is a safe way to store this much energy because a tankful of gas can't be oxidized rapidly enough to do much damage, at least not by accident. Same is true for hydrogen. But compressed air, where the released energy is purely mechanical, is a bit more interesting.
Never mind the extremely high thermodynamic losses associated with compressing (and expanding) that much gas. They're probably on the same order as internal-combustion efficiency, or lack thereof.
Sorry, but the physics just don't work as far as I can see. It really does smell like a scam.
Gasoline is hard to oxidize fast enough, but hydrogen is not.
The LFL/UFL (Lower/Upper flammability limit percentage - the min/max amount of fuel vs air you can have and still burn) for gasoline is 1.4/7.6 - that's a pretty narrow range. But hydrogen is 4/75, so hydrogen will burn in virtually any circumstance.
Hydrogen is, however, lighter than air, so it tends to escape upward before, or while, burning, which limits damage tremendously (except in a tunnel).
Let's be very, very generous and put a cubic meter of compressed air in that car. That would be, rounding up at every step, 30 kWh of energy in the air. If you manage to transfer all of it to your wheels, a small engine might run for an hour on that (http://en.wikipedia.org/wiki/Tata_Nano has a 28 kWh engine) at full speed.
So, with lots of hand waiving, this seems doable, if you are willing to build a 1000 liter tank for this thing, and manage to keep the car and its tank within the Nano's weight. With a more realistic but still IMO very generous 250 liter (about twice the Nano's trunk), you already are at least a factor three of (rounding generously as in every step before). Oh, and did I mention that that Nano will drive about 50 miles in that hour, if driven at full speed, not the 125 claimed?
