I'm kind of an idiot, but here's my understanding: Acetyl-CoA is acetate and Coenzyme A. The Coenzyme breaks sugar down into pyruvate, which can be turned into acetate and attached to the Coenzyme to form acetyl-CoA. In the presence of oxygen, the acetate can be broken down into CO2 and water. If there's no oxygen, the yeast has to turn the acetate into "something else" to get rid of it. The "something else" is yeast byproducts.
It's my understanding that if the yeast is using acetyl-CoA for growth, it can't be used for ester production. I'm not sure exactly how that works, but it squares with info from Neva Parker (WL) and Dr. Cone (Danstar) that high pitching rate = less growth = more esters.
I think there's a strong strain-dependent variable here, because Wyeast, for instance, found that overpitching their weizen strain meant basically no banana esters were formed. So there's a lot of anecdotal evidence that seems to contradict the "scientific" explanation. Although, I think the composition of your malt and your wort should have a pretty big impact on the types of flavors created during fermentation, but hardly anyone talks about that, outside of doing a ferulic acid rest for weizens.
FWIW the best Belgians I've made have been massively overpitched. I've found that using dextrose gives different yeast character than using sucrose or levulose. I used to use a lot of table sugar when making my Belgians, but I've been happier with the results from dextrose. I haven't done any true ABX comparisons to verify, so take that with a sack of salt.