Interesting writeup in favor of brawny cores, by Urs Hölze:

Click to access 36448.pdf

Seems like the big argument is that many operations are latency-oriented, rather than throughput-oriented.  This is clearly true for web processing when a user is waiting for a result, but he makes the point that throughput-oriented batch processing is latency-sensitive, and becomes moreso as you add more cores (because you wait until all jobs to finish).