I agree for typical single-user applications, but it's hard to call that the end of x86 as in the article. If ARM were matching x86 cycle-by-cycle in performance for less power that might be a credible claim that servers were next. Also high-end x86 devices run higher clock speeds with more cores so the gap in absolute performance is at least 1000% percent wide. I just don't see ARM displacing Intel when batteries or at least small form factor aren't involved. I don't expect to have an x86-based phone or tablet either.
My opinion (and I suspect the authors) is that the market for chips where batteries or form factor are not a concern will no longer be large enough to support a company of the size of Intel. RIM makes the finest high security physical keyboard phones anywhere. That is now a market sized for a company 1/10th their current size.