Looks okay for the most part, however, you should consider that the leaving group ability might not come into play. For example, if I reacted a Grignard reagent (Nu-) with formaldehyde I would expect a very rapid reaction since there is essentially no steric hindrance and the carbonyl carbon is quite electron defficient. Even though there is not a good leaving group, formaldehyde is still very, very reactive towards nucleophiles, you just tend to get the alcohol as the product. So, given that, where would you place formaldehyde relative to the other carbonyl compounds? Think more in terms the electron density at the carbonyl carbon. Which other class of compounds would this argument apply to?
Two other points:
- While thio-esters don't exhibit the same level of electron donation that regular esters do, it's not right to say that they don't have any electron donation, because they do have some.
- For acid nitriles, actually CN- is quite a good leaving group. There is a rather famous reagent called Mander's Reagent which has the form (NC)-CO-OMe and is often reacted with enolate anions to make beta-keto esters. What is the leaving group in this case? What does that tell you about the relative leaving group ability of CN- and MeO-?