Early on (around the 10 min point IIRC), he makes a good point about the difference in quality you should reasonably expect for machines that vary in price by a factor of 1000. Our $500 Printrbot is not going to make parts as nice or as large as the half-million professional-grade machine. That is a simple fact. Despite this, I would still argue that our Printrbot is a pretty darn cool capability!
Here's a section I transcribed from later in the video (about the half hour mark) where Wohler is talking about early adopters of additive manufacturing (I took the liberty of adding my links and emphasis to his words).
.. low volume, high complexity, high value parts. [...] If you simply take a design that you have manufactured the old way, whether it's CNC milling or injection molding, die casting, blow modling, whatever process. Just take that design and try to manufacture it with additive manufacturing is possible, but it's difficult from an economic stand-point to make it worth it. The business case just isn't there. You have to in most cases re-design, just as GE Aviation has done, taken about 20 parts and consolidated them into 1 in the CAD software. So to through part consolidation, methods of [...] topology optimization where you can mathematically decide where to put material to optimize the strength to weight ratio. To use mesh and lattic structures, and this part consolidation concept collectively that's where we can see some very big gains. However we don't necessarily have the tools, the tools are very crude, and the work-force out there are still designing for different manufacturing processes. So there's a lot of work that needs to be done, both on the CAD tools and methods of topology optimization and 'light-weighting' as well as creating awareness and understanding the strengths and limitations of these processes...There is a lot to digest in that passage, but I have to agree that there is a huge opportunity for improvement in the area of design software suited to additive manufacturing. There are open source options available for the classic minimum compliance problem, and these use basically the same procedure that more sophisticated and integrated options such as Altair Optistruct use (SIMP or variations thereof).
They talk about "the Maker movement" starting at about 45 minutes in. One of the panelists predicts that multi-material printers will be down at the $500 level (no date given on the prediction though). They expect that "innovation" will be coming from many sources, not just large companies, such as individuals, small start-ups and Universities. GE's GrabCAD contest seems to be an effort along this "crowd-sourced innovation" line that is often associated with 3-D printing. Wohlers mentions that along with the hobbyist ("very low end of the cost spectrum") opportunities for community collaboration there is also some good opportunity for collaborative standards development through bodies like ASTM. He estimates that there are 35k under-$5k machines sold last year (compared to roughly 7k of the industrial grade machines). He argues that these low-end machines are great for people to learn on through first-hand experience, but the average house-hold will probably not be using these machines to make stuff at home.
One of the neat things brought up at the end of the video was a 3-D printed brushless DC motor done by the Keck Center.