Yes, but the reason you don't simply compare the whole string is because not all substrings are equally important. In this case, as per the description, they chose a substring that is known to be important in the formation of functional membranes. Since, evolutionary theory suggests that critical functions are more likely to be conserved than non-critical functions, mutations are less likely to accumulate there, and so you look for correspondences in the important bits of the gene, rather than just throwing the whole thing up and pretending that all the parts are interchangeable.
As an aside, if the rate of correspondences in the important parts do not exceed the rate of correspondences in the less important parts, or if the correspondences in the important parts do not exceed what you would expect from random chance, that would be an excellent way of falsifying the hypothesis that these two share a common ancestor. Or, if they are known by other means to be related, but the correspondences don't fit that pattern, then you'll have blown a big hole in the molecular evidence for evolution itself. Just one more way that evolutionary theory is put to the test every day. But I imagine you already know this... ;)
So 11 amino acids in a range of 50 or so out of a total of about 250 define the "functions" of MOTA?