That paper seems pretty bad to me, doesn’t it?
My read of the evidence is that the genetically-hardcoded snake detector is in the superior colliculus. (But other parts of the brain like amygdala and cerebellum get roped in after within-lifetime learning.) By the way, I think the snake might have to be slithering for superior colliculus to recognize it (and likewise spiders need to be skittering); this is another way that experimental papers in the literature often screw up, I think.