The "early days" myth...
It constantly amazes me that people have such short memories (or for those for whom memory genuinely doesn't extend so far, an inability to research).
It is not "early days" for this sort of technology. Far from it. It has been knocking around in one form or another for 30+ years. That's plenty of time for people to come up with applications - in terms of what the technology can be usefully applied to, and the software to achieve it - while the technology itself caught up.
We have already reached the point where the technology "evolution" is reduced to ever smaller and more convenient iterations of the same thing. Actual progress really isn't occuring any more. But every time a smaller, cheaper or just slightly different implementation of the tech is announced, the terminally amnesiac or plain uninformed declare that "the technology is finally here", it turns out that it still sucks and a new generation of people are born who just think "it's early days" and that the technology just needs to mature.
The tech in question is recognising and tracking in 3D space. Every "new" technology that claims to do this better than ever before in reality is just doing it differently than before. Accuracy reached levels where applications could do something useful with the inputs a LONG time ago, and yet it is issues with precision and accuracy that constantly plague those applications.
This leads the terminally optimistic to believe that the technology still needs to be and can be improved, when in fact the problem lies not in the accuracy with which the input device - the human limb - can be tracked, but in the accuracy with which the operator can control their own device. i.e. their arm.