Yes I actually have recently done tests. I have two systems in the office which I’ve done my test on: dual xeon x5690 3.4 ghz, so two 6 core cpus vs i73930k, non overclocked 3.4 ghz, and yes, the single i7 was 15% (ish) slower then the dual xeon 
If I check their stats on the render farm on the same job, the render times are virtually identical. That is, with xeon having HT enabled, but for this test only, generally I have HT disabled on the xeons because of another ongoing issue we are having with vray.