Paul Appelbaum and colleagues propose four models of informed consent to research that deploys whole genome sequencing and may generate incidental findings. They base their analysis on empirical data that suggests that research participants want to be offered incidental findings and on a normative consensus that researchers incur a duty to offer them. Their models will contribute to the heated policy debate about return of incidental findings. But in my view, they do not ask the foundational question, In the context of genome sequencing, how much work can consent be asked to do?