I recently took advantage of some free time to rework my fantasy setting so that nearly every aspect is explainable through interlocking natural (in my world) laws. Should I have bothered? Should you bother? Not if you don’t care, and most people don’t. I contend, however, that you will design a far more interesting world if you give some thought to the ‘whys’ and ‘hows’ of various fantasy aspects of your fiction. In doing so, you may find that you sidestep common clichés not because you are obviously trying to avoid them, but because those tropes are satisfyingly replaced by the logic of your setting.
A very simple example of this includes fallout from questioning why, for example, your wizard character doesn't wear armour. If it is simply because this runs contrary to the trope, you might consider breaking from convention, but it would seem forced (and leave the reader feeling irked) to dress a mage in plate mail simply to defy the standard. If however, your reader learns that large amounts of nearby magnetic metals interfere with spell casting, then you can dress your wizard in armour made from dragon scales. Now you have a character that is not so typical, and your reader will not feel like they are reading a post-modern commentary on the genre.
Perhaps the more important outcome of this kind of logic exercise is that it gives you a launching point to create unique and flavourful story points. Carrying the previous example further: if magnetic metals interfere with spell casting, then couldn't a wizard be deemed powerless in a room that was lined with loadstone? Maybe all magical devices would be useless? And, perhaps there is a downside to wearing dragon scales. Do they hold a power of their own? Do they act as a beacon in the spirit realm, drawing ghosts, demons, or other dragons? You get the idea. Following the logical extensions of your unique rules opens the way to ideas that will make your story, likewise, unique.
If logical extensions are done well, the reader will appreciate being shown something new. Maybe I am misidentifying the emotion, but it is certainly one of the reasons that I am impressed by serious Sci-fi. Presenting a ‘what if?’ question followed by an answer that is more clever than anything that I could have derived on my own makes me not only admire the author, but allows me to feel safe in the investment that I am making in the story, knowing that I am not going to be let down by some ridiculous, ill-thought plot point that causes me to lose respect for the author and the story. The most egregious example of this kind of betrayal was the script of the TV series, Lost. The story started with a brilliant premise followed by increasingly fantastic subplots that hooked a lot of people. Even many die-hard fans, however, were feeling betrayed by the end when they realised that “lost” simply described the writing team. And midi-chlorians to explain The Force? Even Time magazine openly shat on that one
In the Sci-fi genre, stories that pay close attention to the ramifications of new technologies are placed in the sub-genre, “Hard Sci-fi”. I would like to see more well-done examples of magical systems within “Hard Fantasy” (by which I am referring to the treatment of magic as an interlocking natural science, not to a George Martinesque style of grittiness and adult theme), but they are rare enough that I am having trouble. Brandon Sanderson presents some excellent insights to this topic in his article, Sanderson’s First Law [i], and I am guessing[ii] that his magical systems are among the most coherent. There are several other authors who show concern for hard fantasy, but even one of the best, R. Scott Bakker, would have trouble explaining the full physics behind his magical systems (though the use of different mathematics as an analogy is a good start). Jane Lindskold blogs about the importance of magical systems
, though I have not read enough of her
material to know how well it plays out.
The reason for this scarcity is not due to a lack of good writers. Instead, I suppose that it is for the same reason that nobody (to my knowledge) has done a full wiring diagram of the Death Star. Nobody really cares that much. Fantasy, as the genre name implies, is about losing oneself in an unreal world. Sanderson offers keen insight on this.
 While he is an
advocate of “hard magic” in which the reader understands the rules of the magic
system, he acknowledges that “soft magic” is satisfying when the reader and
protagonist(s) are meant to see magic as an outside force of wonder and not as
an integral part of the plot (as far as the protagonist can affect the plot). But
I still hold my position (and Sanderson’s): if more writers took a “hard” look
at their fantasy, the entire genre could become much richer.
E. Narcisse, "20,000 Per Cell: Why Midi-Chlorians Suck," 10 August 2010. [Online]. Available: http://techland.time.com/2010/08/10/20000-per-cell-why-midi-chlorians-suck/. [Accessed 21 March 2014].
J. Lindskold, "TOR.COM Science fiction. Fantasy. The universe.," Tor.com, 06 January 2009. [Online]. Available: http://www.tor.com/blogs/2008/10/systemunmagical#more. [Accessed 1 June 2014].
B. Sanderson, "Brandon Sanderson," 20 February 2007. [Online]. Available: http://brandonsanderson.com/sandersons-first-law/. [Accessed 13 October 2014].
[i] “Sanderson’s First Law of Magics: An author’s ability to solve conflict with magic is DIRECTLY PROPORTIONAL to how well the reader understands said magic.” 
[ii] I am embarrassed to say that I have not yet read any, but I hope to do so soon (and delete this endnote).