The vast majority of reviews you see aren't going to be done right.
Neither of those reviews are done right, both CPUs aren't generating very much heat at those settings.
And when you see this:
It generally means everything you just read is biased, and that the review site wants more free stuff.
You can make a heatsink review say anything you want, simply by not installing right or messing up your test results. I've seen people recommending an HDT-S1283 over a TRUE because it "gets better performance".
There's no award at the end of this one. Its also testing at a higher TDP. I think you'll soon find out that while the CNPS10x gets alright results around 3GHz, it will quickly fall behind as you approach 4GHz.
Notice how the V10 is noticeably better? Well the Megahalems is even better than the V10, and the Venomous X
should be even better.
....and on an unrelated note I'm going on 11 hours stable at 4GHz now