I don't know if this is an answer really: surely having your right hand on the right, your left hand on the left, up at the top and down at the bottom must confer some advantage?

Perhaps not an answer but alittle background information.

It's thought that the eye went through several stages of evolution in which it started as a flat collection of photo-sensitive cells best able to distinguish between the presence and absence of light but not it's direction. The next major stage seemed to be something akin to the pinhole camera, granting a better resolution (but still pretty awful), true imaging and directional sensing. The stage most like what we have now took that pinhole camera-type design and grew transparent cells over the opening (where we now have the lens). This imparted many advantages in resolution and protection to the eye itself.

The important thing to note here is that as the structure of the eye changed significantly but the arrangement of cells inside the eye largely hasn't changed. Infact the vertebrate eye is still the "wrong" way around and upside down as an artifact of it's origin as an extension of the brain where the structure of the eye grew up around it. In comparson, the eye of the cephalopod is in a more logical orientation due to it's origins on the surface of the head.

So though evolution played a significant role in the development of the structure of the eye but actual arrangement of cells is a holdover from their original orientation and position and little in the way of evolutionary forces has been worked upon them.

All that is true, but I understood the question to refer to the fact that the image on the retina is inverted due to the physics of the lens. A telescope (usually a terrestrial telescope) or a binocular, has ingenious arrangements of prisms to return the image the right way up. The human body, like most cameras, doesn't bother with extra optics, but just flips the image when "looking" at it.

The brain is also clever enough to convert two images into one. (Insect brains - small though they are - can convert hundreds of images into one. It used to be common to see pictures "as an insect sees" them, with repeated images arranged in hexagons; I'm sure this is totally spurious.)

It is likely that the evolutionary advantage is that an eye that focuses an image conveys more survival value than an eye that does not form an image--as long as there's enough brains behind the eye to draw survival behavior from information in the image.  

It's just physics that demands that the image be inverted on the retina, and evolutionary changes picked brain wiring as the best way to interpret the inverted image, than, for example, optical roof prism structures evolving to physically flip the image back over.

Just because the image is upside down doesn't mean (obviously) our brains need bother to bring this constantly to our attention.  If there was some benefit in perceiving the world as our retina sees it, then evolution would have put that pressure on us to see things this way.

I don't think it does flip the image in the sense implied.  It is simply how we learn to interpret stimuli.  A baby's brain learns that light falling on a certain part of the retina relates to objects in a certain direction.  There's no in-born circuitry that "knows" the lower half of the retina, say, is the lower half.

Experiments have shown that even adults can adjust to having the visual fields inverted, and do not instantly adjust back afterwards.