Good and well written article. However, one could stretch your reasoning and also ask the question: market failures... failures for whom? For example positive externalities are at the root of wealth enhancement, and someone out there may want to cash in at this "failure".
But this is not the road I wanted to take; what I really wanted to express, is that assuming asymmetrical information, the whole idea of public policy is getting people to do (or purchase) things they would not do (or purchase) otherwise. Regardless on wether these people are poor or not. You can cast your doubts all you want on the merit of a policy objective (wether it is enrollment, caloric intake, or quantity of beer consumed by a household). In order to have public policy relevance this objective will need to be specified and measured. This does not imply that one objective is superior to another one, but in policy terms you have to have one. And I'm assuming CT are public policy, since CTs carry loses in efficiency always and must be justified for the greater public good.