Using your ear to adjust the levels is something you do in the early stages of the mix.
I understand what you're saying, but have to disagree.
It's the other way round. Meters first -- Ears last.
When the audio goes in you set the clip level against the meters -- the dialogue at about -12db.
Sfx below that about -18 to -24
Music at about the same.
This gives you a rough, but consistent sound level of between -12 db and -9db
Then you decide where the balance between those elements need to change - this is a combination of listening and maths
So, if you decide to up the music to -12, then something else needs to come back by an equal amount.
That's when in FCP you set audio keyframes to manipulate the movement of the levels.
Your final mix is where you listen to the balance between the elements -- and is the part of the mix where listening and having good ears is important.
Truth is that the vast majority of people don't have "good ears" and can't accurately judge the mix -- that's why a sound engineer with "good ears" is worth their weight in gold.
The reason for this is that there are too many element in volved in creating the sound for listening to be effective -- the recorded source levels -- the level of the computer's output -- then the amplification level and finally the speakers.
If you mix by ear, your're only mixing for that set up -- as it is on that day -- and even ambient temperature changes or furniture placement will change the output levels.