Nobody who has ever made a commercial product for a living would be trumpeting a new industrial revolution brought on by open reference designs. Fact is: that stuff is there already. Component manufacturers literally have field application engineers who are just waiting to send you reference designs that use their products, if you're working at the PCB & discrete parts level.

Hardware designs are already open in a way that can't be matched by software - you can look at a product, take it apart, and :voila: - you have the design in a way that is far easier to grok than disassembling a closed binary. Many high end products have full schematic drawings included with the technician's repair manual. Anything novel enough to be patented has mechanical drawings and a full description listed in the patent.

The only things you can't take apart are ASICs and other computer chips. But BFD!? Any player in that space has to have hundreds of thousands just to prototype one, let alone rent time in an IC fab to make them. If you have that kind of money, there's plenty of fab-specific IP floating around for your use at nominal prices.

Open source software is revolutionary because the incremental cost of duplicating bits is nearly zero. The same can't be said for hardware. In fact - it's worse. The smaller a player you are, the more expensive it is to duplicate a design. And it is a pain in the rear to duplicate designs if they're at all pushing the envelope, since manufacturing tolerances and capabilities differ so much from outfit to outfit. Can you imagine the pain and suffering you'd go through porting software that relies heavily on compiler-specific directives? Well, it's like that when you move from one factory to another.

Some Chinese outfits have met with success with open / copycat hardware. I'd argue that's less a result of open reference designs and 3d printing and more of drastically reduced labor and part costs, all situated in a global hub of electronics manufacturing.