I don't have any insight in the test-methodology apart from what the article states, but a few observations make me doubt the relevancy of this test:
The test compares a number of different products: antirootkit scanners and anti-malware scanners. This makes no sense to me. TDSSkiller is an excellent Antirootkit scanner in my opinion, but it is a limited tool, you cannot compare this with a anti-malware scanner like EEK or MBAM because its simply a different product.
The tested malware is for the most part very, very old and not seen in the wild anymore, even though the article states 2015 and "in the wild" in the title.
To give a few examples: Alureon/TDL3/4 hasn't been around "in the wild" for at least 3 years (and thats estimating it very loosely)
The article listed is from 2010 (!) http://contagiodump.blogspot.gr/2011/02/tdss-tdl-4-alureon-32-bit-and-64-bit.html?m=1
The same goes for ZeroAccess/Max++. The latest usermode version of that rootkit was active in 2013 and after the botnet was taken down for a large part, there has been no re-emergence of this malware. However, its kernelmode version was quite a bit older, this was last seen in 2011.
Sure, its interesting to see how products perform against such rootkits, but how useful is it? Those rootkits were "retired" for a very good reason, they can no longer infected today's OS versions.
Finally, I'm not one to make accusations, but I don't like "sponsored by..." tests. I'm fully willing to believe that Zemana was indeed the best product to remove all these infections, but I just think its not the best strategy for any testing lab to let a sponsor also participate in the tests, just to avoid any possible doubt as to the objectiveness of the test results.