I tell my students that in writing, there is not always a precise, narrow, correct way to do something. A sentence can be revised, reworded, reordered in all sorts of way to convey the same idea (but the slight alterations change the nuance just enough). Should your thesis statement have a plan of development? Depends. Should you paraphrase instead of quote? Sometimes. Of course I can give advice on what situations might make a plan of development beneficial, about what sort of scenarios might call for a paraphrase instead of a quote. But because each writer, dealing with unique material, in a unique situation, is going to have to adjust the "rules" or writing to fit his or her subject and meaning, it's difficult to give absolute rules (general rules of quality writing, yes: precise comprehensive rules to explain exactly how to express any particular thought, idea, or information in order to conform to those general rules of quality writing, no). Each writer has to make all sorts of judgment call throughout any paper on how best to convey meaning through language.
Alas, the same seems to go for grading papers. You can use a rubric, attempt to make the evaluation as objective as possible. But when evaluating a paper, you have to make all sorts of judgment calls. How severe is a particular organizational problem? What is the magnitude of this failure to clearly identify a source? You have to measure quality on some sort of scale (though even words like "measure" and "scale" distort how unscientific this can be). Sentences can be awkward, but some sentences are a little awkward, and some are very awkward. Paragraphs can be disorganized, but some paragraphs are a little scattered and some paragraphs lack any clear focus or direction. So, what is the difference between "awkward" and "very awkward" for the paper's grade? How many more points should be deducted for "very disorganized paragraphs" than "slightly disorganized paragraphs"? And what, after all, makes disorganization in a paragraph big or small, anyway?
Unfortunately, I still vex over these things. Though one strives for a fair grading system for each student, I find the best thing I can do is not compare papers. I really have to try evaluate each paper individually, to assess how any writing problems in particular hurt that paper, and try to quantify precisely to what degree particular issues should affect the grade. To be fair to each student, I really must look at each paper somewhat uniquely. I try to evaluate each aspect of the paper (intro, thesis statement, organization, paragraphs, sentences, thesis development, focus, grammar, support, use of secondary sources, proper MLA citations, blah blah blah). I no longer just add up any total from each aspect, but try to bring that all together to evaluate the paper as a whole. But it is a craft, not a science. I look to the department rubric, I try to devise a system, I try to quantify what each aspect of the paper should be worth, I try to determine what constitutes a minor problem or a sever problem, but as a teacher, I have to make all sorts of judgment calls.