Why does a carbonyl reform in ester reactions

I have noticed that frequently whenever esters react with a nucleophile, the nucleophile will attack the carbonyl carbon and eventually in order for the carbonyl to reform the other oxygen will leave for example the first addition of the grignard reagent to an ester: $\ce{PhMgBr + Ph-(C=O)-O-R}$ will result in $\ce{(Ph)2-(C-O^{-})-O-R}$ as an intermediate, and then rather than the $\ce{O^{-}}$ simply getting protonated with water for example, its favorable for the $\ce{O-R}$ to leave and the $\ce{C-O^{-}}$ to become $\ce{C=O}$. Why is the carbonyl reforming more favorable, isn't $\ce{O-R}$ a bad leaving group?

Answer

Alkoxide ion is only a "bad leaving group" in a relative sense (say compared to chloride). In this case, the tetrahedral intermediate resulting from the initial reaction between the Grignard reagent plus the ester is itself an alkoxide ion, and when the OR group leaves, that then becomes an alkoxide ion - so that step should be close to energetically neutral.

Saponification of esters by sodium hydroxide is another situation where alkoxide anion is a leaving group - again because alkoxide and hydroxide ions are close in energy.