Overall, a pretty nice review, saint19.
But to critique you a bit on your testing, you (IMO) need to change up a little on your testing procedure.
1. Do more remounts than only 2. Sometimes it's really hard to get a good mount and I ended up doing 3 remounts as an optimum combination between time required for testing and giving the heatsink a better chance of a good mount.
2. Just use 1 program to stress test the heatsink. Since you are looking for best cooling performance, choose a program that gives you a repeatable and reliable way of insuring max heat production for the test series. I went with Prime95 for my testing because I found it to be very repeatable and can get longer runs in (30 minutes load in my case) without having a falloff in heat during the run. Other programs you might consider using are IBT and OCCT for your stress testing.
3. If your processor will allow it, crank more speed to that 2500k so you can get higher heat production. More heat from the proc tests the heatsinks ability to test hotter procs and is 1 reason why I went with an i7 930 for my testing instead of an 1156/1155 based processor. The 930 is a 130 watt TDP proc whereas the 1155/1156 quad procs are 95 watt TDP procs and are a bit easier to cool when overclocked. Of course you go with what you have, so if you can get 4.7-4.9 out of that 2500k while remaining stable up to 100 C, that would be the way to go. I wouldn't advise going higher on the vcore because you don't want to degrade your chip when testing less than stellar heatsinks.
4. If funds allow, get yourself a pair of high cfm, high static pressure fans to see if a heatsink has some hidden performance in it with a fan upgrade. The Gentle Typhoon AP-30 or AP-31 fans are noisy, but high performance, as are some 120 x 25 mm Delta fans. You can look on eBay to find some high performance 120 x 25 mm server fans also, which can generally be found cheaper than buying them new. Make sure they are matched (same make/model).
And keep up the good work!
