The observation of the elbow jutting out to the side, what I call "floating the elbow," is valid. This does tend to attenuate the power of a punch by taking the arm out of the axis of the punch. Many schools strive to correct this problem by having their students strive to keep the elbow in close to the ribs when punching. Some don't, and we see some pretty flippy punches.
The notion of the forearm bones being in line with the bones of the wrist in the vertical punch and not twisting as they do in the pronated (horizontal) punch is...I think...unsupported. I've heard people say this twisting binds the ligaments of the arms, causes the bones to cross...all this leading to loss of power and structure.
Sounds good, but I'm not sure its true. It'd be testable in part, I suppose, by having people perform a bench press with a standard bar and a modified bar that allows partial supination (I've seen them), using a Smith machine perhaps for added safety. The problem I see with this test is that in a normal bench the elbows are away from the body with the upper arms at right angles to the torso. This is hardly a horizontal punch structure. It'd be neat if someone could device a bar that would allow the lifter to keep the elbows in and do a full or partial pronation of the wrist as the bar was lifted. I can visualize this...but can't imagine anybody going to that extreme.
I also disagree with the notion that one can hit with the full fist when throwing the vertical. One could hit with the full fist with the horizontal as well...but many styles don't advocate this, prefering to put the energy into the index and middle knuckles and sparing damage to the smaller and more flexible bones behind the ring and pinky fingers.
The vertical punch has one very clear advantage in close in punching, of course. That's why people who snap the wrist over in a horizontal punch only do so near the end of the extension. Its a long range punch. The vertical punch can be either long range or short.
Regards,
Steve