This is the real reason here.
Even with this cpu they had multiple ways to beef it up like increasing the clock or use more cores.
Well that's surely another reason. But since we were on the subject of CPUs, let's not act like what the other current consoles have are much more expensive. They may possibly even be cheaper, from a 'per-core' standpoint. So rather than boiling everything down to 'Nintendo pinching pennies again', it's reasonable to think internal development played the biggest role in their CPU choice. Afterall, what's the point of having something a bit more performance driven system if you royally screw your internal teams over in the process?
Seeing the rocky start they had(and described numerous times) in tackling HD development WITH a CPU design and tools they've been comfortable with for over a decade, I'm afraid to even imaging how much more challenging that process would've been with a design and tools that would require some time adjusting to. Yikes... *shudders*
Now that Nintendo has made that HD transition - and with stellar results I might add - maybe they may consider that design switch. But with so many years of familiarity, and tools that matured and evolved over those years, I can't see them throwing that compatibility out the window so easily. They could, but some major restructuring would have to take place internally, and it would - unfortunate - have to begin at a time where Wii U development is finally in high gear(right NOW!), in order for the next transition to go smoothly. Seems like quite a hassle for something that would mostly benefits 3rd-party devs - who won't touch a new Nintendo console anyway.
We're waaay off topic here though, so I think I'll stop here.
(For a while I thought I was in that 'Nintendo seeking lead engineer' thread. This discussion is perfect for that one. )