When I was younger, which I guess is my entire life, I worked at an industrial gas distribution plant. So here’s a couple of observations and an anecdote.
We only pumped oxygen at our facility, vaporized from LOX. But the the plant in Torrance, California pumped everything. Cylinders are inspected before filling, of course, but the worker that day missed a long, narrow but significant scratch down the length of the cylinder. It was later determined to have been pulled out from underneath some heavy object on a job site. So as the pressure neared 2200 psi the cylinder failed. Stress in a cylindrical vessel will make it fail like a banana peel: that is to say the rupture occurs along the long axis. Anyway, the pieces (I think there were three maybe) DID look like pieces of banana peel and were hurled fast enough to kill two workers, on of whom was on the far side of a cinder block wall. When the OP says “a dangerous blast of high pressure gas” that doesn’t begin to cover it!
As for dropping them on concrete, well… Sometime early in my employment we were unloading full cylinders from a truck and placing them on the dock in a loosely spaced pattern. Later we would spin them into a tight arrangement. the way that you get your cylinders off your hand-truck is to grab the caps, tilt the two cylinders away from you, and then hook the hand-truck axle with your foot and pull it away from the cylinders. About halfway through unloading the truck I miscalculated the amount of pull two cylinders could effect on me. I hung on and went over the top of the hand-truck and cylinders and crashed into the stack of previously unloaded cylinders.
Picture a stack of a dozen 150 lb. steel dominoes going down one after the other and you’ve got it. But the worst part was that just beyond our unloaded stack was 24 fully charged and uncapped oxygen cylnders that having just been filled had not had a chance to be capped yet. Most of those went over too. I sat there like a nineteen year old doofus, embarrassed and consigned to picking all of those cylinders up. What I couldn’t figure out was why everyone else on the loading dock was running like hell in every direction away from me. Hell, people in the adjacent front office were running out the front door out to the street!
It was later explained to me the caliber of the bullet I has just dodged. All those cylinders had fallen over domino-style and not a single cracked valve on any of the 20 or so uncapped ones. If you’ve watched Mythbusters you know you can crack the valve off a cylinder with wild, funny results. My co-workers had of course been running away from the wild, funny results.
The upshot of all this is high pressure cylinders are pretty robust. I’d say any of the OP’s falling type scenarios would, at worst, torque the cap and collar assembly off the neck of the cylinder and if the valve cracked off skyrockety effects would ensue.
The “crack/slow leak” mentioned in the last part, on the other hand, is, in my opinion, almost guaranteed to cause an immediate and catastrophic failure with results similar to the above mentioned incident. It’s really hard to put a slow leak (no Scotch tape!) in a fully inflated toy balloon.
