This wouldn't be for TV, so I'll go for how it sounds, in terms of volume.
Therein lies your problem, how loud it sounds in terms of volume entirely depends on your amp/speaker settings and your room. The way it works in film is that you mix in a calibrated dubbing theatre with an experienced re-recording mixer mixing by ear. There are no numbers (beyond the room calibration), levels are controlled by the industry's professionals. I know this answer doesn't help you much but that's the way film sound mixing is done and in general it works very well, unless you can't afford a calibrated dubbing room and a professional re-recording mixer.
As a general rule, little or no compression is used on dialogue in film but this is not an absolute rule, it depends on the production sound, the voice of the actor and the artistic requirements of the dialogue. Limiters are never used on dialogue for film though.
You need to remember that -12db to -6db is not 50% louder; -6db is about 300% louder than -12db (if I remember my exponential maths right).
I think you're getting confused. The decibel scales (all of them) are logarithmic not exponential. As the dBFS scale represents digital values of output voltages, +6dB = double. So, -6dBFS is double the output voltage of -12dBFS. Output voltage is not directly related loudness though, loudness depends on frequency content and other factors, as well as output voltage.
If you're mixing for TV that's another story; you MUST meet the broadcast standards of the station/network. However, you can "get around" volume "problems" by using EQ and dynamics processing.
The new American (ATSC A85) and European (EBU R128) specs make it more difficult to do this. The concept behind these new broadcast standards is to aim more towards the film model, of professionals mixing in calibrated environments by ear. These new standards measure spectrum content and factor in RMS levels, meaning that boosting EQ in the critical band or using much compression will result in higher meter readings and necessitate a volume reduction, which wasn't the case with the older peak level TV specifications.
I get the mixing for sound .. but isn't there some sense in setting a min \ max for your entire film? Maybe I was just raised on TV and expect everything to be pretty much the same volume. ???
As I explained above and on another thread, volume is a human perception, not a property of sound waves, so it's impossible to measure loudness accurately. I don't think TV was ever "pretty much the same volume", which is why new laws were passed in the US last year to try and solve this problem of differing TV broadcast volumes.
G