What a pile of garbage!, This fellow is taking factual data that everyone in the Security Industry already knows and take a totally single sided, sensationalized view and presents it like it's the gospel!
His research and test samples are of a very small subset. If I didn't know better I would believe this garbage and put myself at risk. Very irresponsible journalism.
There are good Malware Products and Bad ones just like any other product we use. That's what the test Labs and Reports are for.
Bollocks, many major antivirus programs now have behavioural threat detection which looks for suspicious program behaviour, not simply signature. Signature-based scanning has been of dubious effectiveness for years, and while it's still a good way to efficiently block known threats, behavioural scanning is what protects you against new ones.
I agree with hotwire_downunder that this research is very sensationalist, though it does have some useful findings; most important is the continuing reliance by anti-virus software on administrator level privileges. We need better operating system support for anti-virus, and by extension anti-virus programs that don't need lofty privileges in order to function; all they really need is the ability to signal to the OS that a program or file may be infected so that it can restrict it, at which point a user can provide administrator support to fix/remove it, the program shouldn't need that kind of access all the time.