^ "Wouldn't a centrifugal machine gun completely sidestep firearms regulations?"
Pending legal analysis by fair-minded legal minds, it looks like it might slip through ATF-type regulations. But (A) many municipalities, for example, define even slingshots as firearms, and (B) so did "bump-fire" stocks.
The main problem with the concept was the (energy ÷ time = power) power to run it. It required a separate bulky and heavy generator --not exactly a "walking fire" proposition.
The US Army once figured it took 300 foot-pounds to produce a disabling wound so that was the approximate minimum energy required for each ball from either the velocity squared or the weight of the projectile, or both. I once calculated the power required to propel a .22 long rifle bullet and it was about 155 horsepower, so there you go right there. *
It's hard to beat nitrates for (energy ÷ time = power) power density unless you go nuclear or pre-compress the propellant fluid or look at percussive materials like azides or styphnates and the like.
This, apart from accuracy problems, but I figure it was one of those "let's try it, who knows, in development, something interesting might show up," as in the magnetic artillery experiments. "And besides, we've got a budget to burn up."
Terry. 230RN
(I'm aware of the magnetic launchers/catapults being tested on aircraft carriers.)
* Check me on that. 1200 f/s, 22 inch barrel, 131 ft lb of energy. Barrel time = 0.00153 seconds, so 85,621 foot pounds per second, divided by 550 to get horsepower, = 155.7 hp. Correct?
But a more likely transit time of 2.5 milliseconds yields 95.3 hp.