Definitely pick up one of the guides. I doubt that choosing one over the other would greatly increase or decrease your score.
Putting in a fair amount of study during the weeks leading up to taking the test is definitely important and can greatly improve your score. Just remember that after you hit a certain number of hours of studying, your improvements will diminish. I wish my adviser from undergraduate had referred me to the specific study (it had slipped his mind), but, as he recalled, students who put in 20 hours worth of studying in the weeks leading up to the test saw significant improvements in their scores. In order to see significant improvement upon that benchmark, students would have to put in some 10 times the amount of hours, and even then the improvements were often rather minimal.
No doubt, the 20-30 hours I put in helped my analytic score drastically because I'm a history major and am uninterested in math, so reviewing refreshed me on what I already knew. On math concepts I was unfamiliar with, I just had to eat the losses (because of the fact that I took the GRE to get into history graduate programs, I wasn't too concerned about that).
With the verbal, you can only do so much to widen your vocabulary, such as the flash card technique, mentioned by Baroque. This, of course, is time intensive, so however much time you wish to spend on it, do what you will. The more important part, in my opinion, is familiarizing yourself with the types of questions that will be presented to you, otherwise you may answer incorrectly questions because of the tricky ways in which they're presented.
For the analytic section: practice writing the essays in the EXACT format laid out in the guides! I consider myself fairly competent at writing and I received a 4.0 out of 6 because I missed some little whatever extrapolation, which the practice guide had mentioned. I recalled it after I finished the exam, cursed myself for it, and the test results verified what I had feared. If I recall correctly, I believe Dak nailed the verbal with a perfect 170, but then got a 4.0 on the essay for a similar reason.
Standardized tests are supposed to be able to divide people on the basis of competency regardless of how much somebody has studied, and indeed they're often effective at doing just that. A friend of mine studied endlessly for the LSAT and could not score well on it no matter the amount of time she devoted to practice. Nonetheless, you can score to the highest of your potential if you're familiar with everything that will be asked of you, and that's where the importance of at least a bare minimum of studying comes in.