Not to knock ARM, but A: I don't know that they have a design for a desktop processor yet (most of their designs seem to be in the Atom/Bobcat realm tops) B: With the absolutely massive amounts of money Intel put's into their Tick-Tock development cadence they have both pretty much the most optimized desktop/laptop architecture their is, and probably the most significant process advantage in the history of semiconductors.
Honestly given the way both Intel and AMD have been able to use out-of-order execution and pipelining to achieve multiple Instructions Per Clock and multi-gigahertz clocks on a CISC-backed-by-microcode architecture I'm not convinced RISC actually has an advantage in practice. In addition Apple is stuck with the foundries, the same as pretty much anybody but IBM, and so pretty much CAN'T begin to produce a chip that will compete with Intel's best when comes to raw performance or performance-per-watt.
For those reasons this would be pretty foolish any time in the next several years. Even if a decade from now they can work past it they will still be stuck fighting off the suspicion that they don't have the advantage they claim to, the one that more or less was true at the end of their use of PowerPC chips.