I'm a wannabe audiophile. And I think a lot of it is a kind of OCD; we want it to be perfect, and if we believe it IS perfect, then we can relax and enjoy it more.
I'm suspicious of 'premium' sound systems enhancing the bass to give an extra depth which makes the speakers sound better on first listen but is actually not an accurate reproduction. One reason why I hate trying to judge speakers straight up. It might sound better now, but is it genuine?
I've heard that CDs are purposely flattened so they don't clip on low to medium speakers so it sounds better on average speakers.
People who work in professional studios tell me that the final mix sounds a lot better before it's turned into CD format but I haven't gotten to confirm this.
Also I listen to the start of Alison Krauss's New Favourite to quickly judge a new sound system on how it holds bass. That song is much harder on speakers than dance music.
-Fred Nerd
You touch on a lot of things I enjoy thinking about. Audiophiles in my experience tend to listen at music, music lovers listen to music. A music lover can get lost listening to the most crackly, fuzzy sounding 78 (old time vinyl for anyone who doesn't recognize that) if the performance is good, whereas many audiophiles will run from the sound. In addition, I've always wondered at audiophile mentality where someone spends tens of thousands of dollars on gear to reproduce music when it was originally recorded on a mixing board utilizing $.29 cent patch connector wires from Radio Shack! An attorney friend of mine who paid $40,000 for a pair of speakers in the early 1990s keeps one of those wires on his desk to remind himself of how his hobby cost him more than it probably should have. And he still owns the speakers which are taller than he is and are kept in a tuned listening room.
It is true that mastering engineers master to the medium - something mastered for radio will sound different than something mastered for mp3 players, and something mastered for CD will sound different yet. But most of us save audio files in digital format. Some use FLAC, I prefer wav as do all of the engineers I work with. There is a reason programs like ExactAudioCopy use wav as the default copying format.
It is also true that what we hear in the studio doesn't sound the same as the finished product, but there are many reasons for that. In our studio we prefer to record to tape then transfer to the computer for editing. There is huge debate over whether analog is better than digital, and I work in both, but I can affirm that after listening to music for 4-5 hours, something that was tracked digitally exhausts me whereas something tracked to tape doesn't fatigue me at all. I also get to listen to first generation sound, meaning it has only passed through recording devices on the way in, and playback devices on the way out, one time. No audiophile has ever heard anything sound as good through their ridiculously expensive systems. But you can only do so much with first generation sound - as soon as you add a second/third/fourth instrument or voice you have to start creating space to hear it, which requires adding eq, delay, panning, reverb, whatever it takes. That first generation sound fast becomes a squashed, artificial sound when individual instruments/voices are lifted from the mix and listened to solo.
There really is no right answer to all of this, except that if you like it it is right for you. I'm a music lover so I don't get caught up in how it sounds as much as I do the quality of the performance when I'm listening privately, but I do my best to make it sound exactly as you want it to sound because I'm delivering a service.