My guess is that most commenters on HN have a CS background and are not immersed in the world of mechanical design. This demo is not unlike that saying about sufficiently advanced science looking like magic. Demos like these are definitely of the kind that provoke lots excitement from people who don't work in the field and, generally speaking, exactly the opposite reaction from those who do.
I must also note that Musk is a deity on HN and, therefore, anyone posting anything counter to what he says is often met with negative reactions. In general terms one is labeled a hater and things go downhill from there.
I straddle the software, electronics and mechanical design worlds and have devoted a significant amount of time and effort to mastering the tools of each discipline. In mechanical design my tool of choice these days is Solidworks. I have also used AutoCAD and ProE extensively.
I would never, ever, in a million years, want to spend eight or more hours a day moving my hands around in the air to design mechanical parts. It would be a horrible productivity sapping experience. A trackball and keyboard, once you've learned the interface, form an effortless man-machine interface to design parts with software such as Solidworks. My preference is for trackballs. When using one, in a very real sense, you are touching, grabbing and interacting with the objects on your screen. It feels very natural. More importantly, it is fast, accurate and ergonomically sustainable for long hours.
I realize I need to provide context for someone who comes from pure CS to get a grip on what I am saying here. Here it is: Imagine someone came along and suggested that instead of using VIM, your favorite editor or IDE you consider using this new interface that requires you having your hands in the air all day to manipulate graphical objects that, when linked together, create programs.
It would look very slick. Perhaps they'd even make a reference to a similar concept see in a movie. You would probably recoil at the very idea and post negative comments on HN. Those outside the CS world would see it as an incredible potential advance in the field and call you haters for not praising an attempt to advance the art.
That's exactly what we are looking at here. I can't think of many people who do MCAD for a living who would think this is actually useful at any level except for one: It might be cool to have as a presentation tool. It would be neat to allow anyone to feel like they are touching and manipulating objects in free space in an intuitive fashion. That's it.
The other more subjective point I'd like to make is that I find myself wondering why Musk/Tesla/SpaceX are burning clock cycles working on anything outside of their core missions. The fact that he had engineers spend time on the Hyperloop and this doesn't sit well with me. I could be wrong, of course, and I am sure there are plenty who will point that out. I just don't think these companies are mature enough to engage in such pointless distractions.