Atoms in a molecule vibrate, yes. In that sense, a helium atom can't vibrate because it's inert - it does not form a bond to any other atom, so there's no restoring force. Think of two balls on a spring vibrating - and now break the spring.
Atomic nuclei probably have some internal modes of vibration as well, but as I'm not a nuclear physicist, I'll plead ignorance on that topic.
In any case, at the quantum scale, these vibrational energy levels are quantized; they can only have specific, discrete values. Your atom in a molecule (for instance) can only reach higher and higher vibrational energy levels through resonance if the difference in energy between one level and the next is the same throughout - true only for a harmonic oscillator. But atoms in a molecule (and likely nucleons in a nuclei) don't vibrate harmonically. At higher energy levels, it becomes anharmonic.
So to use the old analogy: With your wine-glass it's like pushing a swing, you keep pushing at the same moment and the swing goes higher and higher, gaining more energy (until the glass breaks), because it's classical and essentially harmonic (or at least, close enough). Quantum mechanics doesn't allow for 'almost' resonant frequencies; the match has to be exact (or near exact, but that's another story)
To continue the analogy, the situation with an atom in a molecule is like a (somewhat fictional) "anharmonic swing" where, if you keep pushing the swing at regular intervals, it will first go higher, but as it does so, it'll get more and more out of sync with your pushes. Soon enough, you won't be transferring any more energy to the swing. For an atom in a molecule, this will occur well before the bond breaks.
You can still break the bond if it gets enough vibrational energy, it's just that you can't pump it to that level by repeatedly hitting it with its fundamental frequency.