The last, but I think most important point in creating a culture of "distributed R&D" is that you should NEVER punish failure, and ONLY reward on achievements.
This means that if you fail to deliver something during your R&D time, it does not count against you as far as performance reviews. On the other hand, if you do well, it will certainly add points during your review. Every company offering "free R&D time" should embrace failure as an integral and necessary part of inspiring innovation. If your people aren't "failing" (at least marginally or occasionally), they probably aren't taking enough risks.
And if your developers are like ours, or like those on many of the greatest teams out there, they probably have a very similar attitude to most founders/CEOs/serial-entrepreneurs/whatever. In order words, they are likely to be highly entrepreneurial in spirit and are fueled by the taking the risks to venture into the unknown in order to bring breakthrough ideas to life.
That's what's important about failing to succeed during R&D time--because the one great idea that evolves from all of the mistakes or what some companies might see as "wasted" time has the potential to change e-v-e-r-y-t-h-i-n-g.