The problem with "Open Source" hardware, and any other tangible thing, is simply that for most hardware of any significance, a person would need a factory and expensive resources handy to go about trying to make it. Granted, the barriers to ACTUALLY utilizing your rights to modify, update, and redistribute open source software are similarly insurmountable for most people,but this is even more so.
Don't get me wrong, if what this means is that design documentation (schematics, blueprints, manufacturing instructions, etc. etc. etc.) are released with the hardware (so that other companies can use it as a base) is made available, that's still great. That means that Linux will run spectacularly with "Open Source Hardware" underneath because writing the necessary drivers, etc. should be trivial, but I have a feeling adoption will be significantly lower than the open source methodology has been for software.
Underlying it all is the problem of money. Where open source software can afford (until it reaches some critical mass, at which point monetization through advertising and support tends to become a practicality) not to make any money, a provider of open source hardware has to expend significant manufacturing, R&D and Production costs, and most companies won't be willing to simply give away the fruits of all of that effort since the number of people who can contribute back will be relatively limited by comparison (contributors would need to be able to manufacture the hardware to test their modifications thoroughly). Unlike open source software, where there are many contributors, open source hardware would have comparatively few, so the cost to each contributor is much higher and the benefit of having extra eyes looking at the designs much lower. I'd like to be proven wrong, but even looking at the "success story" over at the Make blog, it looks like the vast majority of the "open source hardware" projects were toys with blinking lights and pointless gadgets. Things that might make a fun weekend project, but nothing like what OpenMoko is (was?) trying to do, or that can significantly improve our computing infrastructure and get rid of the problems caused by closed hardware (especially things like video cards, which are still giving open source OS's trouble)