1. Gorilla arm -- keep your hands low. We support tracking and interactions literally 1cm above the keyboard / desk. We're mounting the camera above the monitor to achieve this.
2. We use gestures with built-in physical feedback. For instance, our click mechanism is a "pinch" which brings the thumb and index finger tips together. You can "feel" the physical touch event between your fingers when you trigger a command.
Shameless plug: an old video showing interaction with Reddit, Google Maps, browser. http://www.youtube.com/watch?v=U0WLh7WNxCI
We got one of these devices. And did the demos waved around our hands to move through google earth but then we were wondering so our hands are getting tired and we couldn't really see how this was better than a mouse or joystick.
Maybe it would make it cool for interactive kiosks or cool little showroom gimicks for prolong use, forgetaboutit.
I'm using a Leap controller right now, with BTT for Mac/Touchless for gesture-based control. As I read through a page I can simply stick my hand out and wave it up to scroll down the page - it's a phenomenal experience for passive reading as I don't have to break focus to reach out for my mouse/trackpad. I've also configured some additional coarse gestures to launch mission control etc.
Using the Leap for such brief, coarse gestures avoids both the problems you've mentioned because my arm is resting on my desk, with fingers just a few inches above my trackpad/keyboard, so no "gorilla arm" problems; the gestures are coarse, requiring very little hand-eye co-ordination and finally, the gestures are brief so no fatigue problems.
All of this breaks down once you start trying to do any finer-control gestures, like trying to point at links and click on them like the OP tried to do. IMO the Leap should be used to augment the keyboard/mouse as a secondary interaction interface that you use occasionally.
I totally accept that commenting on a LEAP with out first using one could / will make me look stupid.
The real value proposition is not necessarily in applications designed under existing UI paradigms like scrolling text. It does, however, let you accomplish "science fiction" effects like changing the camera position on a 3D model far, far more easily than the mouse.
I can see this being an awesome technology in an operating room where they don't want to touch things for sanitation reasons. I have a hard time imagining how it is useful in my day to day use of a computer.
Again, I'm not saying that using the Leap for OS interaction is ground-breaking and akin to the first time I ever used a mouse - it's simply a nice addition to my current setup.
1. First and foremost, gorilla arm.[1] My presumption with the "interface of the future" is that it's needed for prolonged use. So, first thing's first, the interface can't be one where our arms require our hands to be higher than our elbows. Unless of course our species got a whole lot stronger in the forearm to support such a feature. Don't see our species doing that anytime soon.
2. Feedback - Right now the feedback loop is eye->brain->hand->brain->eye (repeat) where the hand's pressure against a solid surface is the most important feedback response. With the minority report style interface we currently have a massive delay (comparatively speaking) between the brain->hand->brain loop. We also have to iterate the whole loop much more because we need to constantly assess with our eye where our hand is in 3D (not digital) space. Now let's say the technology gets much better and reduces this to 5ms. We are now bound by the differences of our synapses firing between touch and light. I could be wrong, but it's my assumption that due to the speed of light being the way that it is, that "touch" will always beat "sight" in performance.
For prolonged used applications my bet is on adaptive surfaces. For short term (turning an stove on, flicking a light switch, etc) interfaces I potentially see this Minority Report style interface happening. But does the benefit cost of innovation? Personally I think we are fooling ourselves.
[0] - http://www.ted.com/talks/john_underkoffler_drive_3d_data_wit...
[1] - http://en.wikipedia.org/wiki/Touchscreen#.22Gorilla_arm.22