[Borek]

Think about an orbital as a "charge cloud" - it interacts with another similar "charge clouds". The way these charges interact depends on the cloud shape - different shapes mean different interaction and different properties of the bonds.

[Corribus]

Ok, what abaut my previus question, if orbitals are "just" spaces with some possibility of electron finding, how can they have such apllications in bonds stability, interactions, reactions

Because those properties you mention are also all based on probability and averages.  In the macro-world, those probabilities are very predictable, and so we see deterministic behavior.  Not so at the quantum world, where events are rolls of the dice and completely unpredictable.  Physical rules (such as orbitals) just tell us how likely certain faces on the dice are likely to show up.

[Sossy]

A simple way to describe the distribution of electrons in 1s and 2s atomic orbitals would be as follows: Most of the time an electron in a 2s orbital will be found in a region further from the nucleus than an electron in a 1s orbital. This region is not normally occupied by an electron in 1s orbital. Occasionally; however, the electron in the 2s orbital will be near the nucleus, a region normally occupied by an electron in a 1s orbital. Similarly, occasionally an electron in the 1s orbital occupies a region far from the nucleus normally occupied by an electron in the 2s orbital.The same is true for an electron in a 2p, 3p, orbitals; that is, an electron in these orbitals is usually further away from the nucleus, but occasionally it occupies a region normally occupied by a 1s electron. Would you then say, a 1s orbital is inside a 3p orbital? As you see, the solid model of orbitals is misleading.