Given the fact that, say, for example, at a symphony, the difference between the loudest and quietest parts of a piece are so incredibly massive, are we doomed to fail in trying to create real-world dynamics in the productions we make? what techniques apart from compression can we employ to help us get closer to the ideal?
A truth that we have to take into consideration is that the majority of our work is rarely listened to in a "controlled" environment such as a symphonic music hall. As a TV mixer I know that I have to battle with the real-world dishwasher, a/c unit, neighbor vacuuming, children playing, etc. that barrage the listener as real-world ambience. If they keep reaching for the remote to turn up whispered dialog or calm, creepy ambience over the lawn mower outside, one of those reaches they'll just hit the power button, or worse, the channel + or -.
Now that doesn't win the point that we should all be abandoning dynamics completely and succumb to the loudness war. But I have accepted that "real-world" dynamics rarely fit broadcast QC standards.
However, to get to your question, I have to agree with Justin above and say that improper use of compression (and brick-wall limiting) is one of the biggest culprits that destroys the dynamics of a mix. But this is not new news, and I bet that if we sound editors/mixers were the ones footing the bill on our productions, every single one of them would have a far greater dynamic range and thus a greater impact on our audience. So, while being respectful, I try and educate my producers/directors/EPs every chance I get about the adverse effects compressing the living bejesus out of every single sound has on their story and try to get every dB of dynamics I can get away with. Naturally, I don't win all the time. But it's not for lack of trying.
Sadly it's the broadcasters who hold sway over a lot of the problem. The specs for dialogue for Discovery/PBS is an incredibly small window to get through (4dB if memory serves).
Other specs, such as BBC, are more lenient. I can mix a show with all the dialogue bouncing just under 4 on the PPM and it'll pass, the same goes for 5 1/2 which is 6dB louder. This amount of dynamic range is perfectly acceptable for television mixing, it's wide enough for one to use dynamics, but also small enough that a viewer will be able to enjoy the show at home.
This bring another point surely? To have such a huge dynamic range the room in which the listener/viewer is based would require that room to be treated and quiet. This isn't practical when you don't live in a mansion in the middle of nowhere.
Obviously this only applies to television mixing.
I personally haven't noticed a problem with any cinema mix apart from the very occasional issue (although it's mainly due to bad reproduction in the cinema). My first example I can give is there is a gunshot in Batman Begins that actually hurt my ears. The second was when my wife and I watched St Trinnian's 2. I know it was on DVD at home but I have never had to adjust the volume of my Receiver during a film as the dialogue was too quiet and the musical interludes were far too loud.
Bad use of dynamic range will ruin a film just as much, if not more so, than its lack of use.
I'd say that compression, and over use of it, is the problem not the solution.
The original intention of digital sound was to give greater dynamic range than say a cassette, an LP, or in the film world an optical track, but sadly this dynamic range isn't used all that much these days.
The same applies to film mixing. If we want to get closer to that ideal we have to fight the loudness war.
Sound is a very subjective medium, and people habituate to different aural environments pretty quickly. So, as others have mentioned, one of the most important things a mixer can do is to create a dynamic mix. I think, regardless of comparison to a symphony and other more dynamic events, a good mixer can create a soundtrack that's perceived as being just as dynamic (as long as the listening conditions aren't terrible).
Personally, i love it when a quiet passage in a film has me mentally leaning forward, straining my ears. It's a fine line, but if it's done right, it can really pull you in to a film. Sound is a lot more primal than vision, and so, less critically perceived by the brain (for a typical audience member, and even any of us if we're drawn into the story). I think it's more about perception of dynamics than any objective measurement.