I apologise, I'm not experienced with these things, and cannot interpret these results thoroughly. A previous .JPG within this domain was infected with executable code. Dependant on what engine reading this .JPG, the machine might or might not execute it, thus being infected.
But does this effectually mean that Google search results (and syndicated feeds) may infect a client, if they retrieve the image file directly from the malware site (ie. not creating a thumbnail or equivalent of said image)?