What I mean is that no matter how good the motion sensing hardware, or how good the game coding, the Square Cube Law means that moving our arms mean moving a lot more mass than just pressing buttons, which just moves the thumbs and fingers, or even wiggling joysticks, which involve smaller motions with the arms and wrists.

So again, no matter how good the controller or game code, there will be some lag compared to traditional controllers, and also a greater risk of repetitive motion injuries.

Now workable motion control gaming is still young (not counting non working controls of course, like Mattel's Power Glove), so this is likely something to learn from. Developers seem to be learning that you can't just swap any button press with a motion, just those that the gamer isn't likely to use as much or as rapidly. But games like Red Steel 2 show that even trying to imitate something like swordplay can't be done like actual swordplay, or even the flashy Hollywood Errol Flynn style swordplay. Plus even prop swords will be heavier than a Wiimote or Move. So only someone really practiced can move that fast in real life, and again, people aren't practiced are more likely to sprain their muscles first.

So though it seems to go against a lot of what is learned in gaming, motion control should actually slow it down. Note that in Wii Sports, Boxing requires the fastest motions, and is also noted as the least accurate and more likely to cause strain.

Or with a swordfighting game, focus on accuracy of movement, in either attacking or parrying, to win the battles. And if someone insists on flailing, do like Anthony Hopkins in The Mask of Zorro and have the opponent just knock the sword out of that knave's hand.

Any other thoughts and ideas for how to make this work?

I'm not sure I agree with much of this.  Have you ever played RE4 on the Wii?  It is MUCH faster to aim and shoot than using analog sticks.  As long as the motion controls are used to be an extension of the way you naturally do something, it should make gaming much faster.  The square cube law just doesn't apply here.  Not sure why you think it does.

Moving analog sticks with your finger to aim something is definitely NOT faster than just pointing at it naturally.  If developers start forcing things into motions, something they may have to do with Kinect, then you may have a point.

Why not just use standard controls. Problem solved.

Ignoring the millions of people who've shown the like motion controls IS the problem that too many developers have had. So acting like just dropping this is a solution just shows you are part of that problem.

I like good control schemes so I'm part of the problem? Interesting.

Ignoring those that like the immersion of motion controls is what I meant. Plus motion can be good.

I've yet to see motion control do a single genre better than any existing form of control. Not one.

As I've mentioned, this isn't something that developers are used to. So it just means so far it hasn't worked.

So don't dismiss the entire thing just because the results aren't up to the level of something established in gaming for decades.

How much longer do you think they will need? Another five years? Ten? Are we just supposed to accept that games will suck until developers catch up to a level of control that already exists with KB M or a gamepad?

Why don't think think it could happen sooner, especially with all three first parties now invested in their own motion tech?