All this "minimalist approach" stuff is good...however, if the OP is after a huge prodcution with lots of various instruments, "less is more" is not the answer. It all depends on the genre as well as what you're going for.
If "less is more" is what you're after, you need big sounding instrument sources that are also clear or have a distinct quality about them. Let's rip that Black Velvet tune apart.
Drums: We have a drum sound that is big and makes a statement. Casio drums or even a real drum kit without the impact these drums have in this tune will degrade it instantly. So a big sounding kit with a good forceful kick along with a snare that has an identity like the one used is where we need to be. Don't want to go big with the drums, they must have impact. Listen to that kick drum...this sticks out instantly and gives the tune personality. Listen to the snare...same thing...impact, effects, the whole 9.
Acoustic guitar: not any guitar sound will do. Notice how metalic the acoustic guitar sounds in this. Slight string buzz as well as mic positioning that really make this acoustic stand out. Again, this particular guitar has an identity that would be hard to replace. It's not just a mic'd acoustic...it screams identity and makes the song. Though it is not "big" for what it does, it doesn't have to be.
Electric clean guitar: Really nice bell-like texture that also sound metalic. It also has a slight keyboard type texture to it. Awesome compliment to the haunting keys.
Electric dirty guitar: Any basic blues electric with a strat or tele would work here in my opinion. Nothing to see here other than the tasty licks and solo that were delivered. All that counts of course, but to me this was not a selling point to the song other than the solo and little riffs here and there. The acoustic riffs were more important in my opinion.
Bass: So far we have established an identity with the drums and the acoustic guitar with the electric providing some nice textures. From here we throw a bass in the mix, but it's not just any bass...it's a fretless. Again...another identity instrument. A fretless has a voice like no other. Though the low drone it keep for the majority of the song is the most dominant, listen close to the sound of that drone. It would sound different if that bass had frets. It's also a really nice bass sound and when it does riff, the fretless reason for the bass comes to life to add more identity to the song.
Vocals: Need I mention how incredible this girl is? She could sing with just an acoustic backing her in this song and still deliver it with as much impact. She sells it with incredible vocal control in a bluesey style that just screams that she's lived the blues.
Keys: About as important as the electric guitar in my opinion....good filler, adds dimension and texture and enough of a haunting effect that you can feel the ghosts present within the tune.
In a nutshell, the kick, snare, acoustic and vocals make the song in my opinion. However, the other instruments shine due to their identity. The song is a work of art due to the instrumentation identities that were chosen for it. Without them, it's a boring blues tune with a good singer really. The kick and snare are not even dynamic. Samples that are exactly the same through the song with 0 velocity changes. As a matter of fact, that kick and snare are common samples in quite a few drum modules. But they have impact in this song.
That acoustic is to die for as a lead instrument with those riffs. That bass tone rules and if you don't use a fretless, it changes the sound completely. This to me is a classic case of producing an all-star cast of fine instrumentation...yet all the instruments stick out like sore thumbs in a good way which is what gives the song impact.
For something with more production/instruments, you'd need to re-think your instrument strategy. You can get away with bigger and "more impact" for situations such as this. For something more sonic and rock or with a more orchestrated background of instruments or loads of back up vocals, you have to take an entirely different approach.
Another thing to keep in mind with older recorded music compared to what we have today is, they did NOT use the bass people use today in their mixes. We have way more subs going on which can totally clutter a mix if you're not careful. That's the first thing I hear when I listen to the older stuff.
It's more clear because it's not mixed to focus on all these subs we have today. So with huge productions, you also want to consider how much or how little you push your low end. In most of the 60's music, you heard and felt the bass in it. In the 70's, you started to hear more of the bass than feel it. In the 80's, you heard the bass more than you felt it. In the 90's and beyond, we started getting into those low tunings and more subs entered the picture as well as more "sonics" within mixes....especially with layered guitars all over the place with octaves of the main lines as well as chord inversions to further orchestrate the material.
So my first rule of thumb when creating projects that are going to be huge is to think about how this production will need to be delivered. Do I take the 70's stance and listen to Kansas or YES since they always had lots going on....do I check out some of the 80's pop or rock productions that also had lots going on, or do I look deeper into Trans Siberian Orchestra or some of the good pop stuff like Daughtry and go for that type of production?
You have to look deep inside yourself and figure out what you want to do with your production and then sort of search for a band that may have done something you want to add into your stuff....or at least base your starting point using this band as your template. The best way is to examine what they did and listen over and over while taking notes. If you want deeper low end with more of a sonic sound, you need to check out more current stuff. If you want the middle of the road where everything was sort of balanced, go 80's. If you want a level playing field with warmth and the ability to hear your bass as well as feel it, 70's production is where you should be.
Remember, the low end makes or breaks a project. If you have a common ground with your low end and you're not pushing mids to the point of congestion, you should be able to fit as many instruments as possible in a mix without a problem. As soon as several start to dominate a frequency area, you're capped until you fix that area.
Another important tip...just about every engineer I know goes for too much of a finished sound. What do I mean by that? If you heard what a real mix from a major label sounded like BEFORE it was mastered, you'd be quite shocked. Because most of us are one-man operations, we do it all. But the mistake lots of people make is they try to make their mixes sound like mastered material and they over-do it. The key to your mix and production should be a level playing field...then when you master, you should just be able to polish things up and add all the right stuff. Byt the time most people master, they already have too much sub low and congestive mids in the picture.
Just try some of these things I'm saying...they honestly will make a world of difference. Try to create a balanced mix that doesn't favor anything or lash out with lows. Then, when you master it, add some of the things you think the mix is missing a little at a time. Don't rely on reference material too much...you're trying to mix to something that's been mastered. Get the level playing field first, then add or subtract as necessary. It can really improve the over-all sound by leaps and bounds. :)
-Danny