Sean, An interesting note, but there is no such thing as common externalities today. You claim that externalities exist based on the observation that:
In today's Internet the sender's provider keeps all the money.
How so? In Todays Internet there are various models where the sender's provider keeps all the money, where the receiver's provider keeps all the money, where both keep all the money and where neither get any money!
The sending provider wants the traffic off the sender's network as quickly as possible and onto the destination's provider network.
Not necessarily - there are also reasons why a provider may elect to do best exit rather than nearest exit. What you are postulating is that in a uniform retail model various engineering and financial interprovider arrangements become logically common choices for all players, in so far as the common choice offers the maximal benefit to all players, and then you extrapolate this model into address allocation policies which will optimally service that environment. The weak point in all this is that without a uniform retail model there is not much left where you can generalise about what 'common benefit' may be and what addressing structure will support it. Not that I want to advocate uniform retail pricing structures - my point is only that without such uniformity of retail structures such 'follow the money' arguments as Sean presents here quickly break down to a maze of twisty little passages. Geoff