I suspect the core reason is that there is almost no way to come up with a meaningful number. Only Dragon and Cygnus have a chance of having real hard numbers, since they are actual commercial contracts, that are a pay for service.
Whereas ATV, HTV, Progress, and Shuttle are all government/national programs so it is next to impossible to reliably decide what should be included in their costs. In the case of Russia they are fairly opaque on real costs anyway. Consider how hard it was to get a real/meaningful cost for a shuttle launch.
Simplistically, for the initial CRS contract, NASA tendered a request for 20,000K to the station by each of the two bidders. Cygnus at 1.2 billion/20,000K is about 60 thousand dollars a kilo. Dragon at 1.6 billion for 20,000K would be around 80,000 dollars a kilo. But of course, CRS was also meant to pay back for some of the development of the systems. (Even so, 2.8 billion for 40,000K of cargo plus two complete launch systems and cargo vessels is a literal steal. That is probably less than one year of the SLS budget, for nothing real till 2021)
So the contract AFTER CRS will be the one of interest. Once SpaceX/Orbital have recouped development costs through a lucrative first contract, how much will the second services contract be worth? Presumably a lot less. (One hopes!)