Thanks for the feedback @localdeity. I agree that my article could be read as implying that emotions are inherently irrational and that, from an evolutionary perspective, emotions have underlying logics (for instance anger likely exists to ensure that we enforce our social boundaries against transgression). This reading does not reflect my views however.

My scheme follows decision theory by assuming that we can assign an "objective" utility value to each action/option. This utility value should encompass everything - including whatever benefits may be reflected in the logics underlying emotions. Thus, there shouldn't be any benefit that emotions provide that is not included in these utility values. There are times when our emotions are aligned with those actions that maximize expected utility, but this is not guaranteed.  Whenever an emotion goads us to act in line with utility maximization we can call that emotion "rational." When the emotion spurs us to act in a way that conflicts with our best interest (all things considered), we can call that emotion "irrational".

My goal in this article was not to argue that emotions are fundamentally irrational. Emotions operate according to their own internal rules. These rules are more akin to pattern-response than to the sorts of calculations prescribed by decision theory. This article tries to integrate these effects into decision theory to create a model of human behavior that is qualitatively more accurate.