Likewise, even if there was some trait that improved a species in some way toward perfection, such as in humans, super-vision or immunity from cancer (temporarily assuming both of these would be possible), unless it was strongly associated with the ability to pass down one's genes, it would not be favored by natural selection; these super genes would never become widespread because the individuals with them wouldn't out-compete their peers. This dependence on improving reproductive fitness of a species completely limits the ability of natural selection to even theoretically produce our ideas of perfection.
Another thing to consider is traits that are both disadvantageous and recessive, such as recessive genetic diseases in particular. Even if they are detrimental to a species and reduce an individual's reproductive fitness, they tend to stay in existence because heterozygous individuals (that is, those with both a dominant, in this example advantageous copy of a gene and a recessive, in this example disadvantageous copy of a gene), when the dominant allele "hides" the recessive, might not display any signs of the disease or even ever be aware that they have the recessive allele. Because of this, their reproductive fitness will tend to be unaffected, and because only those who inherit two of the recessive gene would tend to die without reproducing, the gene would perpetually remain in the species.
While natural selection does not fashion idealistic perfection, it does manage to "give" species certain traits that can be considered "perfected" for their environment. Complicating the matter of a "perfect species" though, is that the optimal traits will vary over time depending on inevitable environmental changes. The most successful life forms tend to be the ones that can adapt to the environment over generations through natural selection the most effectively, but as described earlier, the mechanisms in place for species to change over time will not lead to perfection.