Executing on this idea as successfully as they have is far from easy.

To my untrained eye, it seems like "raise lots of money, buy many, many patents, build nothing so the target can't bargain with their patent portfolio, and then start negotiating". Once you have that machine working you can iron out the details. It's like domainers - get there first, buy everything that looks vaguely useful and extract rents.

What's the hard part?

It's no harder than "build a faddy social network thing that's slightly better than MySpace", or "build a slightly better search engine than AltaVista" to other untrained eyes, I guess.