There is definitely room for improvement for next-gen hardware, but it is a difficult time for introducing it right now. There are too many uncertain factors that make it very difficult to decide what kind of investments and trade-offs need to be made.
There is:
- 3D Controls
- 3D Display
- a tonne of different online services present and/or emerging
- OS and embedded developments related to the above online services (think Android, Windows Phone OS, etc., with their related software distribution platforms)
3D tvs are being introduced this year. Move is coming this year, and Natal likely as well. I think it will take two years to see how most of these developments will trend out. Taking that into account, I think either Sony and Microsoft both design two completely different consoles, focussed on two different approaches, and then at the end of those two years decide to run with either the one or the other.
Finally, publishers don't seem to be dying to get new hardware to play with, as the new multi-threaded rendering pipelines and dx11 haven't settled yet in the PC space, and consoles still have a little room for growth and experimentation (though limits are definitely being hit, maybe a little more on the 360 which perhaps is reaching its maximum potential a little sooner).
Another tough decision is whether or not next-gen will be a good time for massive multicore. That is going to be a tough choice, but an interesting one, especially in the context of that the importance of running cool and being cheap may be bigger than ever. Certainly I want to bet that Sony is going to set an upper limit for itself of 399$ out of the gate, with good prospects on going down to 299$ after 1-2 years, max.
However, I do think it shouldn't be more than two years before at least the publishers get their hands on new hardware and/or tools. Perhaps this time it could be a good idea to get publishers in on the new stuff a little sooner and allow them to work with finished hardware for a year before the actual consoles are released.
For both Microsoft and Sony, I think an important part of the strategy will be to determine what to do with the existing userbase. If Microsoft wants to approach the casual space with Natal, I don't think they can get away with killing off the 360 in the way they did with the original Xbox at the release of the 360. But running two services at the same time isn't great either.
This isn't much different for Sony. Presumably, they will both be upgrading their service and upgrade the existing services on the old consoles so that the current gen consoles run on the same network with limited functionality. It shouldn't be too big of a problem considering that they will both be having a platform that spans from handheld to console, to phone and even TV, but it may still be complicated - what if Sony wants to go Android, for instance? Can they afford to put Android on the PS3 as well as the PS4? Most likely it would be much cheaper to do that on the PS4 with new and more powerful hardware, less backward compatibility to consider perhaps, and so on.
Above all though it is not unlikely that the hardware will be the least consideration. This is the lesson that Wii gave them, and Apple is giving them all over again. They will need to think about what they want to provide their consumers with, and what they want to sell them, and design the hardware for that, rather than just design really cool hardware and see what cool things third parties will do with it.