There comes a point in studies when it stops being useful to read more and more, and begins to be useful to write all the time. A two-year period is about right, as it turns out, to learn enough about a field's many facets to transition from exploration to contribution. It's very important to know when that point has come.
A good literature review ought to be a chance to point to important research that leads to the story you're about to tell in your paper---your inspirations and motivations. It's not a good time to start exploring related but irrelevant work. The content of the story you tell is the crucial aspect, and I'm seeing now that the best papers need not cite very much up front if their story stands alone.
Writing is a habit, and notes can be taken in paragraph form with inline citations just as easily as they can be bullet points. The slow train of non-experimental academia proceeds as each individual explores, learns, and consolidates their own particular preoccupations into a new perspective. There's a major risk that nobody will in fact climb upward---whatever that means---but the hope is that cumulative semi-shoulder standing will raise to greater heights.
But me, I want products and gee-whiz-look-at-thats. By the way, look at this word and tell me why it's so remarkable:
More on the way! Right now, packing for Boston and an Echo Nest internship. Then, taking the Statistics Q1s in September and powering through to a 5-quarter MS Statistics. Onward!