Robin Hanson may very well be the clearest and deepest thinker on the planet, so it's no small problem when he blogs "Why I'm Not Libertarian". Let's see if we can rescue his inner libertarian.
Public goods: intellectual property. Hanson claims: "You have no fundamental right to enjoy the innovations produced by others without compensating them." A libertarian would counter: you have no fundamental right to give innovations away and then demand compensation from those who chose to enjoy them at your initial offering price. If you can use contracts or technology to propagate your innovations without actually giving them away, libertarians should not call you a force initiator. Your only problem will be when someone breaks your contract and shows the innovation to a third party not bound by the contract. This problem smells like it could be solved through some combination of insurance, bonding, and watermarking. If (like many libertarians) Hanson worries that such a solution might not be efficient, then he should at most support the patent value tax and copyright that exempts non-commercial use.
Common goods: natural resources. Hanson writes "It is probably sometimes efficient to initially allocate property in other ways than via the usual 'making'", and links to his interesting article about orbits and sunlight. We geolibertarians can only shout "amen" to his point about the unjustness and inefficiency of the "royal libertarian" notion that first use/control of a space (land, orbit, spectrum, air corridor) creates the same sort of property right as in an object. As for the general case of non-excludable goods (a.k.a. common goods a.k.a. natural resources), the Lockean mandate to leave "as much and as good" translates quite naturally into pollution/congestion taxes and exclusivity auctions. The same sort of analysis when applied to land yields a land value "tax".
Creation of persons. Hanson writes "It is probably efficient to endow parents with partial ownership of their children." Maybe, but you don't need formal legal title over an agent you can program, and that's effectively what we already do to our kids through genes, mammalian bonding, and acculturation. This should be all the more true in Hanson's future filled with artificial persons mass-produced via emulations of very-carefully-chosen template brains. Hanson doesn't make the case that libertarian principles are necessarily inefficient here.
Social contracts. The most unlibertarian statement Hanson makes is almost an afterthought: "And it is probably efficient to enforce non-explicit contracts, such as among very large groups." One could take this as a placeholder for all the usual debates between anarcholibertarianism and minarcholibertarianism. I prefer instead to cut this Gordian Knot with the magic of radical Foldvarian federalism, where the leaf nodes are essentially homeowners' associations, and secessionists are dared to fully opt out of the services and protections of their community. As far as I can tell, Foldvary's geolibertarian vision could deliver all of the efficiencies Hanson might desire from social contracts.
Harm vs offense. Hanson writes "people can hurt each other 'non-physically' via info in so many ways". Hanson doesn't seem to recognize how slippery a slope he's on here. He wants to talk about information (gossip, blackmail, etc.) as being dangerous/harmful, but any "info harm" here is fundamentally a case of not liking how somebody else has used his freedom of association. If you can claim a "harm" just because other people are associating in a way that is suboptimal for your utility, then every act (and omission) is "harmful" to pretty nearly everyone. Freedom of association is efficient, and I don't see Hanson carving out a coherent set of "info harm" exceptions.
Conclusion. Hanson is right to demur from Bryan Caplan's allodial anarcholibertarianism. Most of Hanson's complaints can be addressed by the principles and prescriptions of geolibertarianism. However, he might only call himself "geolibertarian-leaning", since he has made it very clear that his root value is economic efficiency rather than personal liberty. That's OK -- we don't need to debate whether geolibertarianism is good because it works, or works because it's good.