Isn't instantly scalable access to system resources what everyone had been working towards for decades? Isn't that what many shops had in-house with their Z's?
Amazon was forced to create a scalable infrastructure to support their primary business model. In working all that out it just makes sense for them to spend a bit more and sell their excess capacity, which is what they are doing. I'm all for it!
The rubber will really hit the road when/if some truly HPC applications utilize the system. If the fog can truly scale with security there will be no reason for specialty firms like Cray and no place for big boxes from IBM.
In truth I'd buy foggy infrastructure from Cray or IBM for HPC because I have faith in their backend. They don't mess around. I'm not sure a retailer (Amazon) has a vested interest in making sure my trial models really stand up to scrutiny from an engineering perspective.
Crog? Zoud? How would you name a cloud/fog from a proven vendor. Maybe that's why they don't do it... Maybe they should, if AWS really works; big iron in-house (the big money) might not have a place in a few years.