yes, the questions in those books are arranged by (putative) difficulty level. as far as i know, there are no concrete statistics concerning the 'actual difficulty level' of the questions; however, many of the questions are actual material from previous official GMATs, so we can safely assume that GMAC collected student performance data for the questions.

from my anecdotal experience working with students, 'difficulty' is a much more concrete issue on math problems than on verbal problems. in other words, it's much easier to decide that certain math problems are 'hard' and others are 'easier'; on the other hand, a verbal problem that's 'hard' for student X may well be 'easy' for student Y.

here's some functional advice, then: you can probably get away with segregating your math study by difficulty, but you should probably take the time to go through all the verbal questions (as there will almost certainly be some, even at the beginning, that are randomly 'hard' for you).