I would disagree on a sprawl being like rooting, because a sprawl is not a static thing, it involves lots of adjusting, moving, fighting, and other elements. Simply sprawling and holding will have little effect on stopping a takedown, the other person will just adjust there angle and put you on your back.
However you are right about the single line of attack can be positioned against. Simple body position and knowing what angles things are strong / weak at. And it is the problem with the demo's like the one above, the "attackers" only do one thing, at one angle, after the defender sets his position.
Anyone that spends time doing submission grappling will learn that positioning is very important, a difference of a inch or two often means the difference of whether a lock can be finished, or will be escaped. This is the same effect as seen in the video above. The defender had good positioning, the attackers weak. As they did not adjust or change there angles, they where not able to complete the attack.
It's not rooting, or chi based, or anything fancy, its just body alignment (which I believe is claimed) and is something well known in sport arts as well. The big difference is that it is not a staticlly trained or used skill.
So lets try this, grappling theory 101...
New people, almost always have difficulty applying locks to the same effectiveness as the more advanced people. Regardless of art, when the new guy tries a lock he is often met with blank stares. When the instructor does it, it feels like your arm is going to get torn off.
Yet to the new guy he is doing the exact same thing, but without the pain. Minor adjustments make the difference. a couple degrees here, half a inch there and "OUCH!" it starts working. Over the years people fine tune these angles and positioning, making the locks more effective with less effort. I think its safe to say that goes for everything from sport grappling to Aikido.
So a very minor adjustment is what makes a technique go from ineffective, to effective. The flipside also works, a minor adjustment on the defenders part can do he same. This is why joints need to be isolated and controlled to apply a lock, you need to be able to control the opponents ability to free himself.
Now, just as you get better at adjusting to make things hurt more, you can get better at learning to adjust to make them hurt less. If you train it, some people don't, they only train in static situations with the only thing getting covered is applying it, not escaping it.
So, overtime, it even becomes a little instinctual, you pick up that certain angles make things hurt, and others don't. So when attacking you try to get to the ones that do, and avoid the ones that don't.
So far, this all goes fine for the video above. He has merely positioned himself in such a way that the angles and leverage points of the attackers are useless against his posture.
However, nothing is static, and that is where strategy and adaptability come in. Rarely do you get to try one technique attacking one angle and have it work. There is usually a chain of things to set it up, follow ups when one thing doesn't work, etc.
To take a traditional example, something I remember seeing as a "aikido arm wrestle," start of as a normal armwrestle, switch directions trapping the other persons hand with your other hand. Again, grappling theory 101, don't attack strong angles, attack weak ones. Force him to react one way and use it against him.
Same works in striking, strike at a low target to open up a high one. So an important thing in defending is to protect against more then just one angle, and not over react. If I drop both hands to block a low kick, I may very well stop the kick perfectly. However the follow up punch to the chin is going to hurt a lot.
However most of the rooting examples I have seen are exactly that. yes, it stops a attack on that angle, but opens up a ton of other things. Basically the equivelant to blocking a low kick with both hands, yes, it does stop the kick, but it is still a bad idea.
"Rooting" may work very well in isolation, but that doesn't make it a good tactic. It will fall apart when the isolation is removed and the other person begins adjusting to there defences, they will essentially stop the low kick but get KOed for exposing there face.