I think Yahtzee is right, he's just not *completely* right. The current generation of motion controls are cheap gimmicks. That doesn't mean that all motion control are cheap gimmicks.
As I type this, I am using Firefox as my internet browser. I have an add-on called 'FireGestures' (which is one of several such extensions), allowing me to use my mouse as a gestural interface. A quick swipe to the left to go back, a quick swipe to the right to go forward, an L-shape to close a window, and so on. It's quick and easy.
Many people have observed that sweeping gestures and whole-body movements make playing the game more work. And I agree with that assessment. However, the concept of a gestural interface does not require flailing around.
Let's go back a few years, to when the Wii was new. Commercials for Wii Sports show the players swinging the controller like a bat or a golf club. But when I actually played the game, I quickly discovered that you don't need to do that - and in fact, it's not even the most effective way to play. A short, quick flick of the wrist achieves a result that's as good or better than the flailing full-body movements.
And as far as I'm concerned, that's where the potential for motion control lies - the ability to make quick flicks that are no more taxing than pressing a button, and the ability to point the controller at the screen and use it as a 3d mouse. Also, like MovieBob, I think the best part of the Wii control scheme is the ability to have an on-screen pointer (which is actually better, in my opinion, than either a computer mouse or a touchscreen - it moves where you point, without your hand blocking your view.)
The ideal I hope we can work towards is something like a Kinect, but able to track eye and finger movements rather than (or in addition to) the full body movements. If you can move your finger an inch, or even a few inches, and give a command, then that's not a appreciably more work than moving your thumb from one button to another. Point at the screen and have a cursor (or targeting crosshairs) appear where you point. Give commands with simple gestures - for example, to pause the game you hold up your hand, palm outwards, in the universal signal for 'stop', rather than adopting a full-body pose as the Kinect currently requires.
There will always be a place for sticks and buttons, simply because interface elements like 'push forward on the stick to move forward' or 'hold down the button to keep firing' are not as easily replicated by motion control.
(Incidentally, this is why I think the Wii's 'nunchuck' controller is still a vastly superior motion-control device, even years later - it combines a fairly traditional thumbstick, d-pad, face and trigger buttons with a point-and-click interface and simple motion control for both hands. The Move is more heavily focused on the motion control, and the Kinect is entirely reliant on it.)
Anyway, where was I?
Ah, yes. I think that (perhaps barring the sort of Direct Neural Interface which Yahtzee alludes to) there will always be a place for traditional controls; however, if developers are willing to move past the 'standing up and acting out the motions because that is what looks good on the TV advertisements' gimmick, and acknowledge that most of the time you want to enter commands with as few movements as possible, Motion Control can and should have a place right alongside them.