Minacom seems to have conducted a more thorough study whereas the Brix results were strictly based on results of tests run by users of its web site. Those results are skewed because there is no way to know if the users of the Test Your VoIP site were only those having problems. It wasn't a random sampling of call quality across geographies.
These results are certainly good news for VoIP providers (though they do note that the PC-based VoIP services (e.g. Skype, Google Talk, etc.) did exhibit worse quality than the VoIP over broadband services.