About all three
Here's the reason all of this is so confusing - the tone of an instrument is essentially a mathematical problem that can't be solved analytically. The tone of a piece of wood all by itself is a problem that can't be solved analytically. When you add multiple pieces of wood (body, neck fretboard), other materials like plastics etc....you get something that's even more of a mess. Finally you add the electrical component which is also imperfect.
The biggest factor in the sound of the guitar is of course the electronics. This is the signal that's going directly in the amp, the zeroth order. Next are the main mechanical components, body, neck, bridge etc, first order effects. These indirectly affect the sound (by how they feedback with the string). Now the effect of the finish is essentially a second or third order effect (as in unimportant if the finish is a small fraction of the mass/thickness of the guitar). This is because the main change is in the acoustic sound of the instrument. The acoustic sound then has to travel through air affect the strings, which is a very weak feedback given the loudness of the instrument (as opposed to the direct mechanical coupling of a bridge or the electrical effect of pickups). So as long as the finish's thickness is a small fraction of the thickness of the instrument, the effect will be negligible. The degree to which it is negligible will of course depend on
the thickness.
For what it's worth, the imperfection of wood is what makes it tonally interesting. A very homogeneous material (think a piece of metal) has a very sharp
frequency response centered at its resonant frequency. What happens when you tap a piece of metal? It rings at one pitch right? In mathematics, that tap closely approximates a signal with evenly distributed frequency spectrum (that is it has the same amplitude across all frequencies), so what you hear after the tap is the "response" of the material - you are hearing its EQ curve. So the metal is strong in a very small frequency band. Wood is very much different - it is organic and chaotic. There are knots of higher and lower densities, pores, etc, that each contribute to give a much more complex frequency response. Now you know why luthiers and builders "tap" pieces of wood - it's a quick way to approximately determine the frequency response of a material whose frequency response is very hard to predict. This is also the reason different instruments of the same make/model/year/color sound different.