I would respond initially by saying that the question is somewhat malformed. In what sense does any literary product have a purpose except that provided by the author, which is necessarily individual? Even if the author defines a purpose, should that have any bearing on whether the text is perceived as having that defined purpose?
I personally subscribe to the view that in matters of interpretation, intent is irrelevant. What the author meant to do, insofar as we can even know it, has no bearing on how the work can or should be perceived, in no small part because what a reader perceives is more valid than what the author thought they were creating. Perception is the conversation. I also tend to think that unless we can have universal access to intention, by which we would need not only biographical and personal writings, but also actual access to the mind, then an author's intent is useless to us. How am I supposed to know what the author really intended to do? This is not to suggest that we can't discuss intent, mind; rather, I'm suggesting that we shouldn't assume intent as the sole arbiter of interpretation or perception.
However, purpose is something quite different from intended-reception. whatever the author intended as the purpose of a written work need not determine how we interpret that text’s purpose. Intent and purpose, in other words, are different beasts, as the former concerns the activity of production while the latter merges production and perception together. We can, after all, discuss the success of a text in its presentation of a message while also discussing the other interpretative possibilities of a given text. Indeed, the purpose, insofar as one is defined, only offers possibilities, as it does not suggest "this is the only way to read the text," but rather that "the author meant to do Y, but what we see are A, B, and Q." (Alternatively, it might be helpful to avoid the total linguistic separation and simply make a distinction between "purpose" as an intention" and "purpose" as an end product. But maybe that's abstract, too. Oh well.)
To return to the question of science fiction's purpose: as I noted in my post on the taxonomy of genre, science fiction doesn't seem to me to fall under the traditional category of genre anymore because it lacks the narrative devices which define all of the other market genres (crime, etc.); science fiction, in other words, is a supergenre because it is conceptual, though it s possible to think that at one point, science fiction had a narrative practice. In a similar sense, I think the purpose of science fiction has been obscured by time. At one point, the most obvious purpose for the genre might have been to entertain (as in the Pulp Era) or to expound upon the radically changing world of the 40s, 50s, and 60s, and so on and so forth.
Now, I think the genre's purpose is less apparent, and perhaps for good reason. It can entertain, experiment, extrapolate, examine, elucidate, and encapsulate. There is no singular purpose anymore than there is a singular narrative space. And that's another reason why I think science fiction is one of the most important literary genres, as its narrative spaces, purposes, and perspectives exist in an endless sea of variations. One can write science fiction for any number of reasons -- and one should feel comfortable doing so. Entertainment, experimentation, whatever.
The idea that we can identify a singular or minute number of purposes for this genre is an exercise in futility, because science fiction cannot be a genre of limits if it is to also be a genre of endless narrative possibilities.
What do you all think?