I disagree.
Antivirus may fail one day and another day pass, but say it's useless, it's farfetched !
I did not mean to say your reviews are useless. Rather I am a fanboy of you.
What I wanted to say is, if you judge a AV based on few samples then it is not fully correct. If you use lots of malwares to test it then every time you test, the result will be almost same.
For example.
If I use 2 sample to test one AV and one of them missed, that means it can detect/block 50%.
But if I use 2000 samples to test the same AV and that time 200 missed, then the detection is 90% and that will be a better result. It's just the concept of probability.
And your reviews only proves my point. In one of your review the same AV scored excellent but now it's a horror. The main reason behind that is for both the cases the number of samples was very less.