Uncertainty Principle for Compound Objects Interpretation The most common misstatement of the HUP seems to be that it's a constraint on knowledge of a system or particle, rather than a description of the behavior of the system itself.  E.g. an electron is completely described by the wavefunction in a hydrogen orbital, rather than having a definite position within the orbital that we discover by measurement.
While this seems intuitively easy to grasp when considering electrons and individual particles, I struggle with how to conceptualize this "completely described by the wavefunction" existence for compound objects like protons, nuclei, and molecules that need to maintain strict geometric relations to one another.
Put another way:
I can visualize that the electron IS the wavefunction describing the electron field in the double slit experiment, which is why interference effects are observed.
For a compound object like a molecule that still displays interference effects, I'm finding it difficult to reconcile the uncertainty principle that limits localization of these tightly bound atoms, and yet allows for interference effects.   
How does the HUP allow for the tightly geometrically confined configurations of atoms within molecules, and simultaneously for large compound objects like molecules to be de-localized to the extent they can exhibit interference effects in a double slit setup? 
 A: Suppose you take two weights and put them between two walls, and attach all of them with springs. So it's wall, spring, weight, spring, weight, spring, wall. This system will have two oscillation modes, and these two oscillation modes can be excited independently. If we keep it simple and say that the weights have the same mass, and all the springs have the same spring constant, then the weights could oscillate back and forth in unison (so the distance between them stays the same), or they could oscillate towards and away from each other (so the position of their center of mass stays constant). Every oscillation of the system can be written as a linear combination of these two oscillations. The set of possible oscillations is a vector space, and those two oscillations make up a basis for the space. Thus, we can analyze the system as being an "atom" of two weights, and the atom has an excitation amplitude, and there's an independent excitation amplitude within the "atom"; the two amplitudes are independent and can be analyzed separately.
Similarly, when you have a hydrogen atom, there will be a space of quantum states the system can be in. We can choose a basis for this system in which all the excitations within the atom (that is, the electron orbitals) are elementary vectors. We can then treat the entire system as being a state of the entire atom combined with a state within the atom; those two types of states can be separated into two linearly independent spaces. When we have an entire molecule, the number of internal states increases, but the basic principle remains. I'll continue discussing a hydrogen atom, rather a molecule, since the analysis is simpler
Since a hydrogen atom can be split into external states and internal states, it's quite possible for the position uncertainty to be quite high for the external state, while quite low for the internal state. That is, it's quite possible that we have very little idea where the proton is, and very little idea where the electron is, but still have a very good idea what the distance between them is. Even though the proton is a "blur", and the electron is a "blur", these are not distinct blurs. These are highly correlated blurs; measuring the position of the proton highly localizes the position of the electron, and vice versa. The two particles are entangled; there isn't a position for one particle that's independent of the position of the other. Neither particle has its own wavefunction; the two-particle system has an overall wavefunction, and that wavefunction has a "where is the atom" part and a "where are the two particles in relation to each other" part.
Note that if you had a free electron and a free proton, the two particles would make up a system that would still have an overall wavefunction, but this overall wavefunction could be split into a "electron part" and a "proton part" that would be largely distinct. It's once they form an atom that the wavefunction can  no longer be approximated by treating it as a simple decomposition into two wavefunctions that each deal with only one particle.
A: Nuclei in a molecule do exhibit zero-point motion about their equilibrium positions. The amplitude of this zero-point motion depends on the shape of the potential and on the mass of the nucleus. I believe that at room temperature, most small molecules are in their ground vibrational state, so that the energy is $(1/2)\hbar\omega$. If the potential has a soft shape, this zero-point motion can be large. It can, for example, be large enough to make a metastable molecule spontaneously break up. In ammonia, the molecule can tunnel from one orientation to the other, like flipping an umbrella inside out in the wind, but the tunneling rate is low. In general, visualizing the nuclei as having classical positions and trajectories is a better approximation than visualizing the electrons the same way. This is because of the greater mass.
The same kind of thing happens in nuclear physics. Nuclei are not rigid spheres. They can have a variety of equilibrium shapes, and those shapes can fluctuate. For example, there are fission isomers, which are metastable, highly deformed states of nuclei, which can decay either by fluctuating to a more extreme deformation and fissioning, or by fluctuating to a normal deformation and gamma decaying.
