I got tired of waiting and just built a i5 2500k system, so in about 3 - 4 years AMD will get another shot at my money.
From purely a technology standpoint I'd be really really really curious if AMD continues to invest to stay in the x86 rat-race come 4 yrs from now, or if they seek more promising markets while minimizing their presence in the x86 consumer space much as Via was forced to do.
Consider that 3 yrs from now Intel will be releasing their 2nd generation 3D-gate tech in the form of 14nm.
The whole GloFo foundry idea was based on the premise that the foundry would realize cost-savings by taking on more customers for the nodes, diluting the R&D cost across the volume of multiple customers. The only problem with this vision is that customers, other than AMD, are not interested in high-performance SOI.
I have not asked, things are in an interesting state of flux at the moment, but I would be quite surprised if GloFo produces an SOI option at 22nm, let alone 14nm, the economics for R&D development are just not there (as AMD expected, compelling them to spin-off the fabs in the first place).
That said, looking at what Bobcat was able to deliver on a foundry-based bulk-Si process I can't really see the continued justification for AMD and GloFo to insist on sticking with SOI.