I've been told by many guys that instruments should have their own space.
This involves widening the sound and also panning.
But as far as I can tell from how much Dance Music I've been listening to for the past 2 and a half years,I've yet to hear a professional track with extreme panning or even to an extent that you can really make out that it's favouring one side more than the other.
Take Hardwell's latest single,Blackout
If you haven't heard it already,give it a shot.

As for the drop,I can make out that the leads are widened,but I haven't heard an instrument that's panned.
Can anyone explain,how these professionals are doing their Stereo Imaging and just how exactly are most of the instruments occupying space at the centre?

put 12 humans in a small elevator...

that is basically what your sounds are doing... everything needs its personal space... the leads sit somewhere in the mix and have their own freq's the drums, the bass, the pads etc...

you can use widening plugins, EQ's, and Reverbs to help with that. What I like to do is mix in groups (bucket mixing) so all my sounds are grouped by section (leads, drums, bass, vocals etc...) and then I mute the group channels and bring them up one by one until they sit nicely where they should. I can add master FX to the groups If I want a bit of reverb or something to spice up that section.

I think when people say all sounds need a "space" in the mix, it's usually referring to the frequency spectrum, not the stereo field. width is also important, but it's more important to not have too many elements overlapping in the same frequency range, especially the low-end.

You need actual physical space as well as a place to occupy, in order to have your sounds to take up space and not be crowded.

I do not agree with your bias, but i will say that both are incredibly important if you want to achieve a balance in your mix, other wise you'd literally be having to many things on top of each other but still occupy the same space.

think of it as a box  where people are standing on top of each other. Different frequency ranges, but nothing is being panned because you stacked em up like jenga blocks. This applies even if the dimension are capable to having people more spread out, which you do and that's called panning.

It's a combination of three things which are: frequency, panning (or physical space), and volume (how pronounced you are at filling up your physical and frequency space)

Not to mention width control, which roughly falls into panning and physical occupation of space.

I wonder what program this is... https://www.youtube.com/watch?v=0Bc5TYO_1Ik 

This reminds me of this old video (1990?). He used almost the same graphics to explain mixing.

https://www.youtube.com/watch?v=TEjOdqZFvhY

There are lots of ways to spread a single sound out in stereo:

• Autopanning them so they swoosh across the stereo field
• Using "doubler" plugins that use short delays panned left and right (look out for mono compatibility problems)
• Longer, tempo-synced delays panned more or less hard left and right
• Stereo reverbs... maybe with some mid-side processing to widen the reverb?
• Chorusing (watch out for mono compatibility again)
• Xfer Records/Massive's "Dimension Expander" effect - like a stereo delay/reverb hybrid but designed to be mono compatible
• Using a stereo widening processor on the sound itself (mono compatibility warning!)

Mono compatibility? Well... if you take a sound, and a slightly delayed copy of the sound, that were panned left and right, and you turn the audio into mono... then the two copies of the sound will interfere with each other. Depending on the length of the delay between them, certain frequencies will be boosted and others will cancel out. Worst case, that sound will be degraded or made much more quiet by conversion to mono. Any effect that creates stereo using delay might be susceptible to this kind of problem.

But there are alternatives to just widening one sound.

One trick often used with vocals is to record 2 or 3 takes of a verse or chorus, and pan one hard left and one hard right. ...And maybe a 3rd, placed in the centre. This wins you a load of stereo width, and it's mono compatible, because the sounds aren't exactly the same: they're separate recordings, so there'll be differences in pitch, timing and tone, so if you collapse the stereo mix to mono, they shouldn't cancel each other out significantly.

And there are related tricks you can do with synths and samples:

Maybe a main synth in the middle, with not much energy above 7kHz, plus two different synths (different plugins/wave forms/filters/tunings/modulation... different NOTES?), panned hard left and right, and EQ'd so they have a sizzling top end? Or at least, side-panned synths that are timbrally distinct from the lead in the middle.

Or, maybe rather than one warm, muted pad, you have 2 different ones, again maybe from totally different synths, EQ'd differently, maybe playing different notes, and panned left and right?

Or, instead of one stereo reverb, how about 2 different mono reverbs (1 plate vs 1 spring?) panned left and right?

Or 2 different ride cymbals (an acoustic ride vs a 909 ride?) playing the same pattern but stereo panned?

Also, based on an answer to a question I asked here a few days ago...

I've been using stereo percussion loops (EG a bar of mixed latin percussion, a live congas loop). But I've been experimenting with making them mono, then panning the mono signals. It feels like... maybe 2 different mono'd samples panned apart, sound more spacious than the same 2 stereo samples similarly panned. I guess because the stereo samples overlap more, whereas the mono samples occupy only a slice of the stereo field?

Google 3D mixing, it has a similar philosophy to mixing you might find interesting.