Taste is a subjective thing and it's pointless trying to construct a theory that rigorously or scientifically justifies one's preferences. But that's not going to stop me from having a stab at why bad science fiction is, well, pants.
In an excellent post, Martin of Wis[s]e Words fame discusses how science fiction literature inhabits a second class existence. He talks about the feting of literary authors when they have a dabble in sci-fi (he mentions PD James, but also Margaret Atwood, Kazuo Ishiguro and Philip Roth spring to mind), even if the science fictional aspect is, well, pretty dodgy. And also the assimilation of prominent (but often long dead) SF authors into the literary canon.
To my mind nothing illustrates the dichotomy between "proper" literature and SF more clearly than the prodigious output of Iain (M) Banks. For his literary works he is simply known as Iain Banks, and for the SF the M initial is inserted. As a reader of all his work they are entertaining, clever, original, and witty. None of his SF can be shoe horned into the big-gun-bug-eyed-alien mould, even though there are plenty of them about. But as far as the literary world are concerned, some of his output is more worthy than others. According to the first edition of this book's attempt at creating a new literary canon, only one Iain M Banks makes it into the 1,001-strong list (The Player of Games) while a handful of his literary novels make the grade.
Can this be put down to general snobbery? Probably, but not solely. But what makes literature literary is, crudely and generally put, the adherence, or a commitment to subvert a set of accepted narrative strategies, which usually prize style, plotting, subtle invocations of influences and contemporary fashions, and characterisation. In other words, literary authors work in a particular field encompassing other authors, publishers, lay and academic apparatuses of criticism, and audiences with certain expectations. Ishiguro's Never Let Me Go can get away with it not just because it is by a respected name, but because it conforms to certain literary conventions. Iain M. Banks's SF, a superb author like Peter F. Hamilton, at least for now, fall outside of these conventions.
The field of science fiction operates in much the same way hence why it is easy to dismiss literary forays into its domain as not 'proper' SF. For example it is quite possible to write a book or series of books around one or several big ideas and still win sci-fi plaudits because value is attached to speculation over and above traditional literary value. Authors like Richard K. Morgan, Stephen Baxter and Harry Turtledove are certainly in this mode.
Following Bourdieu, fields do not exist in isolation. They inter-mesh and overlap other related and formally unrelated fields to varying degrees. Similarly everyone simultaneously inhabits multiple sets of fields and sub-fields. Some are in complementary relationships while others are conflictual and contradictory. In my own case, as a regular reader of science and literary fiction, I am aware of the "capitals", strategies and stakes each field possesses. Among others I inhabit a sociological field, circumscribed to a degree by specialism, geographical location, and academic biography/career; and the socialist end of the political field - again conditioned by my party membership, the issues that particularly interest me, and my experience as an activist. A theory of taste, or rather more accurately, a theory that explains why I like what I like can be constructed from concepts generated from this framework.
Let's illustrate this with a novel I've recently read. John Birmingham's Weapons of Choice is the first in an alternate history trilogy that sees a multinational naval taskforce from 2021 accidentally transported across time and space to the mid-Pacific on the eve of the Battle of Midway. The rest of the book is spent with all sides - the task force and the 1942 combatants trying to come to terms with the "transition", with a touch of 21st century meets WWII scrapping along the way.
The premise of pitting modern day technology against the 2nd World War is not a new one - Harry Turtledove did it over four turgid books in his World War series (where alien lizards from outer space have a go at invading Earth in 1942). And there are a couple of obvious winks in Weapons toward it. But what makes it objectionable to me is not just the cardboard characters and unlikely plotting (seriously, in a middle of a war and unsure about how much 21st century technology Japan had salvaged, would precious and irreplaceable materiel be expended in POW rescue missions?), but also the drawing of the female characters - who are either Paris Hilton alikes who do guns and kung fu, or dowdy lesbians - is awful, and the overall sociological naivete really grates. It offers an apologia for the War on Terror (accepting the thoroughly discredited claim this is a clash of civilisations and ideals) and puts a liberal gloss on the allied involvement in WWII, which, of course, was all about stopping the bad guys. Birmingham really does believe it too - there's more than one occasion when characters from the future lecture their forebears on the theory that history is merely a clash of contending ideas.
Weapons fits nicely into the alternate history/techno thriller sub-field of SF, but you cannot help but bring the baggage of other fields to bear when forming a judgment. It almost completely eschews literary values, sticks in the craw of my political values, and on top of that does as much violence to the laws of social development as the premise does to the laws of physics. But, again showing up the subjective limits of a sociological theory of bad sci-fi, Weapons is not intended as an intervention in any of these fields, but as a cultural artefact it cannot help but be interpreted through the filter of other fields and judged according to tastes informed by them.
Bourdieu's concepts provide a method for investigating literary tastes in much greater depth than this illustrative example, but it cannot arbitrate on the question of value. It is a way of explaining why some things are valued and others are not. What it cannot be is the final word on good and bad taste.
(An alternate and more formal way of spotting bad fiction is here, courtesy of Feminist SF).