As a student I was heavily obsessed with Mahler (as were many of my contemporaries, probably due to the efforts of Leonard Bernstein to promote Mahler's work). The only real problem was that his obsession was fed almost entirely by recordings and occasional broadcasts of live performances on the radio. (Unless I am mistaken, there was a Public Television broadcast of Bernstein conducting Mahler's 8th symphony; but it was on when I was still having a lot of trouble getting my mind around that work. The first movement of the 8th was also part of the inaugural program for Lincoln Center in what, at the time, was called Philharmonic Hall; and that concert was telecast.)
Now I live in San Francisco, and Michael Tilson Thomas has made this one of the best cities in the world for hearing live performances of Mahler. (I also recently discovered, thanks to my RSS feed from the London Telegraph, that Thomas is doing the same thing in London with the London Symphony Orchestra, where he is a "permanent guest conductor.") What I have learned from these experiences is that there is always more in a Mahler score than can be captured by any recording process, no matter how good the technology is; but, because there is so much complexity for the ear to navigate, the recordings can provide a sort of "outline," without which many of the subtleties of a live performance might be missed. They also allow one to hear any single work of Mahler's with some sense of his entire corpus (or at least a healthy portion of that corpus); and this, too, can be informative.
For example, consider the fact that Thomas just conducted the Mahler 4th in London. Lots of people who write about music, particularly for the general public, like to claim that this is his most "accessible" symphony, particularly since it comes on the heels of the 3rd symphony, which is one of the most challenging for any conductor, not to mention the audience. (The first movement of the 3rd is longer than most symphonies. I remember hearing a radio broadcast of Mehta leading the New York Philharmonic. In spite of the fact that, by that time, clapping between movements was one of the great cardinal sins of the concert hall, there was a solid, not to mention well-deserved, round of applause after he had steered the Philharmonic through that first movement.)
The "accessibility" of the 4th may have to do with the fact that it can be heard as a reaction to the 3rd. Indeed, the final movement of the 4th, a setting of the Wunderhorn song, "Das himmlische Leben," was originally intended as the final movement of the 3rd, with the programmatic title "What the Child Tells Me." (As a matter of clarifying my sources, most of my understanding of Mahler comes from the de La Grange biography.) I think Mahler himself recognized that the scale of the 3rd was already getting out of hand (even if much of his reputation at that time was for conducting Wagner); so he dropped the "Child" song.
When he then picked it up again for his 4th, things started to get interesting; since one of the inner motifs of this song setting became the seed for the plan of the first movement of the 4th. Then, in the orchestral eruption at the climax of the third movement, we hear the first suggestion of the principal motif of the song setting. Whether or not this tightly-knit structure makes the whole symphony more "accessible" is debatable; but it encourages us to hear the symphony as an integrated whole in its own right, rather than material left over from the 3rd.
Once we start playing this game, however, we discover that the arrow of time points to the future, as well as the past. If we go back to the first movement of the 4th, I believe we shall find that the trumpet fanfare at the moment of its climax, just before we settle comfortably into the recapitulation, is pretty much a note-for-note forecast of the fanfare that introduces the funeral march at the start of the 5th! Once we start doing this sort of thing, we discover that it is actually pretty easy to play the game of finding threads that link the symphonies, particularly if you include Das Lied von der Erde, whose final ("ewig") motif moves over to the harp motif that introduces the 9th. Then, of course, you have the Lieder Eines Fahrenden Gesellen preparing material for the 1st and Wunderhorn Lieder motifs showing up in the 3rd.
Am I suggesting that, from a logical point of view, Mahler wrote only one composition, which occupied most of the years of his practicing as a mature musician? If that is the case, then it might be better to view the Mahler corpus as a kind of diary. We know from all sorts of evidence (the comments scrawled over the pages of the 10th being the most blatant) that Mahler endowed his work with a lot of personal involvement; so this diary hypothesis is not particularly far-fetched. Music history provides us with plenty of examples of composers reflecting on personal incidents, such as the departure of Bach's brother or Beethoven's rage over his lost penny. Wagner certainly let a lot hang out, particularly when he was willing to surface his reflections on an illicit affair in his Wesendonk songs, which, in turn, served in part as a sketchbook for Tristan und Isolde. We now know that a similar theme served as the "secret program" for Berg's "Lyric Suite." However, these are all examples of episodes. Mahler seems to have been driven by his whole life-experience in a far more consistent manner, which is why I believe that, whenever we listen to Mahler, it is valuable to be informed by the whole corpus of his work, even if we can only get at that corpus through the limitations of recording technology.