Easier? How does more complexity equate to being easier? One could assert that you can have greater success through the manipulation of two variables than just a single variable but I am not seeing how this equates to an easier overclock...
Overclocking is simple enough as it is for Sandy/Ivy Bridge; what we want is better overclocking ability/potential.
If Intel had surprise success with improving timing margins to provide 1GHz higher average clock yield, they would simply launch a product lineup with most of that 1GHz factored in.
With multipliers locked across all lower-end models, Intel ensures that overclocking headroom remains relatively modest compared to what it used to be when most low/mid-range chips were still overclockable by at least 33% by simply changing FSB clock ratio.
Maybe I'm missing something, but how is this any better than just using the core multiplier?
It should greatly benefit non-K chips. The biggest winners from this should be the Pentium's & i3's which have no available "K" versions or Turbo Boost and have been virtually frozen at under 3.5GHz for 2 generations now with typically only +/- 5% BCLK to play with... Many people have been crying out for the return of super-overclockable budget dual-cores ever since the i3-530 Clarkdale (with 2.93GHz -> 4.3GHz overclocks) got "upgraded" to the i3-2100 (with 3.1 -> 3.3GHz overclocks) when Sandy Bridge locked "core" and "uncore" frequencies together...
In fact, overlocking on BCLK-locked Sandy & Ivy Bridge i3's has been so poor that 1st-gen "i" series BCLK unlocked 2010-era i3-530's are still faster in many benchmarks than the 2 years newer Ivy-Bridge equivalents. Examples:-
If this BCLK-unlock is "across the board" and not just limited to premium "K" chips, then it's great news for Haswell i3 owners, as a stock 3.4GHz i3 could be OC'd to 4.25GHz (with 1.25x / 125MHz BCLK) and not just limited to 3.5-3.6GHz as current Ivy & Sandy i3's.