You can say that applying a given force someone's head will apply a greater torque and cause a greater rotational acceleration that applying it to, say, the chest, That, though, does not seem like a realistic scenario in gymnastics.
I think the more interesting case considers rotational momentum, which is rotational inertia multiplied by rotational speed. Since rotational inertia depends on mass distribution, "tucking in" while rolling reduces rotational inertia and so increases rotational speed (because rotational momentum is conserved). You spin faster when you curl up and slow down when you spread out.
I think I first considered this learning Aikido. When I want to stop a roll quickly, I uncurl and my rotational rotational reduces (and the floor applies an opposing torque to my arms and legs. Another aspect is that being tall provide a greater lever arm for an opponent and can be easier to know down.