I thought the control oscillations were terrifyingly large, unless they were intentionally showing off. From a safety or design standpoint, there's zero reason it needed to be swinging that far. Kudos to the control system for keeping it upright and landing where they wanted it, but I get the feeling they came within a gnats-ass of crashing.
How far it fell ballistic, and how close to the ground it got before the rocket started braking for the hover and landing also makes me suspect they have a very narrow fuel window, possibly because it had to be that low to make weight, but also cuts into your hover/landing time drastically.
Although I concede that Birdman could pop in here and tell me why I'm wrong in five seconds flat.
Nope, not wrong.
Also, they can get away with it more than spaceX as their rocket is basically an order of magnitude smaller (the whole thing is about as big as an F9 LEG!), so it's terminal velocity is way slower, thus doesn't need as much impulse to stop it).
However, lower length/diameter means harder to balance (think 12" ruler on your hand vs yardstick) and lower lever-arm for the off axis thrust = larger deflections required. Also, single engine vs 9 that all have variable throttle.
The -only- thing they did that wasn't done before is (pardon the Boolean)
((Return from suborbital >20,000ft) ANDNOT (with substantial excess velocity) AND (land successfully) AND (on land) AND (on surface level with surrounding)
However, that's really not that amazing. Both DCX (roughly the same size) and grasshopper (way bigger) did all of the above, just at much lower altitude, and F9 did all but successfully land MULTIPLE times, in a much harder environment, AS PART OF a mission that had nothing to do with landing. SpaceX is -only- doing the landing as a "why not, it's cheap testing on something we are already doing, where we -really- don't care if the landing works, and it doesn't impact the primary mission).
Hell, the only reason grasshopper couldn't do the same altitude as BO is they really didn't care to.