I read that articlet as a general caveat, that if ARM device makers can keep Intel from increasing its process lead, then they will likely manage to use software to overcome weakness in ARM-based hardware.
Well there's risk that earthquakes will gobble up Silicon Valley, too, but you don't see many authors cautioning about that one - even though the likelihood is about as high as having the foundries beat Intel to 14nm with a more power efficient process. In fact, an earthquake may be even more likely, given that the foundries are already set to offer a 14/20nm hybrid process, which is very unlikely to compete on power/performance with Intel's true 14nm process - even if the schedule did come greatly ahead of expectations.
As much as one might like Windows, and the Intel processors that enable it to run, the tablets that do not contain windows and Intel, outsell the Intel-inside tablets by a big margin. So clearly, neither Windows, nor Intel-class processors, are needed in a successful tablet. At least, so far that has been true. How would that change in 2013 or 2014 ?
People often fall into the logical trap of assuming that the market doesn't demand something, just because there is activity without it. While I wouldn't deny that tablets have ramped incredibly fast without Intel, the fact remains that this does not imply the market wouldn't shift to Intel based tablets, once Bay Trail and Haswell arrive with highly competitive metrics.
Market Data 
Markets 
AI
