I think in true what happens is that when you "touch" the two together, you actually aren't - on a macromollecular scale, yes they are touching, but if you consider it on a more mollecular scale, there will be other things coming into play - namely the fact that firstly, there are still going to be large air particles in between them most of the time (I mean, air like 70% nitrogen, and that is a pretty big diatomic mollecule). Even if the air is not in the way, you would still end up with numerous impurities on the surface, which would stop it from working (I think is the case with Aluminium, which forms an oxide layers just a few atoms thick, but which is strong enough to stop it from reacting with many other substances).
Lastly, even if you did manage to get them close enough - well, the electrons are delocalised because of something called orbital overlap, which is how you get any electron sharing. In things with delocalisation, you get a large orbital which shares the electron density throughout. Now, this, when it forms, is stable and has a certain energy level that makes it stable. However, trying to attach another couple of atoms of metal would involve breaking these orbitals, at least on the surface, and reforming them. You can do the reforming thing when you twist or otherwise mess around with metals, because the delocalisation isn't attacked as badly, and hence does not constitute a high energy barrier. But in terms of attaching another molecule, I think the activation energy i.e. energy required to break the previous orbital, is too high to be overcome under normal circumstances.