Its hard to draw definitive conclusions when you're speculating from a sample of one! Imagine showing a child with no knowledge of animals a snake and asking her to describe what she thinks the other animals on earth are like and the habitats they occupy. I think there'd be a risk that she'd describe a range of snakes and possibly lizards, but wouldn't be able to imagine something radically different like a whale/eagle. That's the risk we run when our only sample is the earth.