[Winga]

The resonance effect in acetate or phenoxide are quite different than that of benzene.  The biggest difference is that acetate and phenoxide are anions, which tend to be higher in energy than neutral molecules.

I'm guessing that you mean that benzene is more stable than cyclohexaTRIene.  That is because of the resonance effect, but those are both neutral molecules.  You're kind of comparing apples and oranges.  A closer comparison would be between an acetate anion where all the charge was fixed on one oxygen (no resonance) versus the actual case where the charge can be smeared out over both oxygens.  However, this has no effect on the protonated acid stability.  That O-H bond has to be worth something, right?  The resonance effect of a carboxylic acid just makes the deprotonated form lower in energy.

Sorry, I am really compared the stabilty of benzene with cyclohexaDIene. From data, benzene is more stable than cyclohexadiene.

If I use cyclohexadiene to synthesize benzene by elimination, 2 C-H bonds of cyclohexadiene are needed to break, then, one C-C pi bond is formed. After the C-C pi bond has formed, resonance occurs. Because of this resonance, extra energy is released (we know that bond break require energy, bond form release energy). Resonance energy plus C-C pi bond energy can overcome the energy to break 2 C-H bonds, so benzene is more stable. (Am I correct?)

Back to the case of acetic acid, I think it is difficult to compare the stability of acetic acid and acetate. But, can I say, the neutral compounds must stable than the compounds carrying charge(s).

[movies]

For the benzene thing, comparing benzene to cyclohexadiene isn't really a "fair" comparison because they have different numbers of bonds, right?  Benzene: 6 C-C sigma bonds, 3 C-C pi bonds, 6 C-H bonds.  Cyclohexadiene: 6 C-C sigma bonds, 2 C-C pi bonds, 8 C-H bonds.  If you compare benzene (with resonance) to cyclohexaTRIene (assuming no resonance) then the difference you see is the resonance stabilization energy only, right?  I guess you could do the same calculation from cyclohexadiene, but you would just have to account for a lot more energies in the process.

For purely covalent compounds the neutral form should always be more stable than the charged form.  I can't think of a contradicition of that.  There are some compounds that are stable but have some sort of charge separation within the compound, but overall their charge is 0.