OK, so here is my having slept on it, read the fine print, and really tried to punch holes in it evaluation of the Sealand/Havenco Datahaven.
With the exception of Avi Freedman most of the people involved are not who I would consider first choice. Their two biggest Liabilities teamwise are the CEO who comes in with strong ties to carribean money laundering, and their Chief Logistics officer who comes from the family of nutcases that own this island.
Second, their physical security planning is very sketchy from my view. I have come up with four different ways that they could have their operation shut down with minimal outlay of cash and the action not traceable to the group who planned it. I'm not even a professional spook, they've probably got twenty or more.
In the tradition of their CEO I am posting their odds of success at 1 in 7.
If they do succeed they will have significantly lowered the bar on the operation of DataHavens. This may open up the opportunity for others to compete with them, by providing more secure locations with better management.
If they fail, It will mean that the bar is right where I have thought it was. In that case here is my proposal. an island off the coast of Costa Rica, fiber from the mainland, fiber from asia, and the redundancy of the sattelite uplink(it would probably be a good idea to actually own the satt), a serious crypto corporation, like Zero Knowledge or Counterpane or both, A company with solid credentials in network architecture, and external security handled by some very hard guys.
If anyone with money or motivation agrees with me on this, email me.