I don't agree with this. I think the aldehyde C=O is more polar and more reactive than a ketone. I can understand how your notes may say a methyl group would increase the bond polarity in an effort to explain the boiling point only on bond polarity.
I always like to turn to real data in these kinds of situations and in this case unfortunately the data does not support your conjecture.
Important point: the permanent dipole moment of acetone is 2.91 D and that of propanal is 2.52 D. These are fairly consistent for all ketone and aldehydes, respectively. (For sake of completeness, that of formaldehyde - a "double aldehyde" is 2.29 D, even less polar.) Probably the most likely cause of this is that a ketone has two fairly rich electron donating aliphatic groups, whereas an aldehyde only has one (and formaldehyde zero). True there are an equivalent number of CH
n groups on both acetone and propanal, but really it's the number of CHn groups directly adjacent to the carbonyl that count - i.e., and ethyl and methyl have fairly similar electron donating capacity because the second carbon out is too far away to have much of an inductive effect. Anyway, the long and short of it is that ketones have higher dipole moments, therefore they are more polar, therefore they will tend - all other things being equal - to have higher boiling points (e.g., boiling point of propanal is 46-50 °C and that of acetone is 56 °C). While all the electrostatic and intermolecular forces will play a role in the boiling point, permanent dipole-dipole interactions are far stronger than London or van der Waals forces, and so any minor differences in these secondary effects in acetone vs. propanal are far outweighed by the large difference in permanent dipole moment between these two species. Obviously, if you start comparing propanal to something like 2-octanone, where the sheer number of van der Waals forces starts to become very important, then it's no longer a fair comparison.
I think the reason for the apparent contradiction in the OP's post is because he is forgetting that while boiling point is strictly a function of electrostatic forces, reactivity is a function of electrostatic forces and steric considerations. Ketones simply have less room for reaction, regardless of the electrostatic favorability of a reaction - and formaldehyde, far more reactive than either aldehydes or ketones, have the most room. In addition, the dipole moment mentioned above is for the whole molecule, not just the carbonyl, which can easily become a point of confusion. While the presence of two electron donating groups in a ketone paves the way for a large overall molecular dipole moment, it also appears to reduce the
positive charge located specifically on the carbonyl carbon.* This would again mean that ketones are less reactive than aldehydes, because reactivity relates to electrostatic forces at a specific point, whereas boiling point considers the electrostatic forces over the entire molecule (and in between them).
So, to sum, I don't really see a contradiction here. It's just requires a deeper explanation.
* Now my turn for conjecture. I puzzled over some other data for a while trying to explain why ketones have higher dipole moments than aldehydes. I first turned to vibrational frequencies of carbonyls, and aldehydes typically vibrate at lower frequencies than ketones (1730 wavenumber vs. 1715 wavenumber for propanal and acetone). This might tell us something about the length of the bonds, and hence the polarity, but I came to the realization that it's not helpful because the mass attached to the carbonyl also makes a difference, so that is a confounding factor. I next turned to bond energies, which shouldn't really depend on the masses at either end of the carbonyl, and found that for the three classes of carbonyls, BDEs are typically in the range of ~712, ~733 and ~754 kJ/mol for formaldehyde, aldehyde and ketone, respectively. My supposition would be that this translates into a stronger, shorter bond for a ketone versus a longer, weaker bond for aldehyde (and longest, weakest for formaldehyde), which in turn means less polar carbonyl in ketones and more polar carbonyl in aldehydes. Or at least, more pi-bond character in ketone and less in aldehydes. Contradiction, you might say! I did, too. But after thinking about it a while, maybe not. It is, after all compatible with the idea of extra electron density being fed into the carbonyl carbon of ketones. This lowers the polarity of the carbonyl itself (because the carbon is less positive) but increases the overall dipole moment because that positive charge has to go somewhere: it is distributed (to some degree) onto the electron donating methyl groups (in the case of acetone). Dipole moment, after all, is related both to the magnitude of the positive and negative charges - and their degree of separation. So in the end, orgopete, you are probably right that aldehydes have more polar carbonyls, but ketones, overall, have larger dipole moments, because the charge is separated over large distances. That's a hand-wavy argument on my part that may not explain everything (if indeed it explains anything - probably it is vulnerable to assault by a determined individual), which is why I left it as a footnote.