The Genomics Law Report asked that question in a recent roundtable of researchers, bioethicists, and lawyers. In his commentary, Daniel MacArthur, a scientist with the Wellcome Trust Sanger Institute, asserts:
… withholding medically useful (or even simply intellectually interesting) information from research subjects even if they request it – is ethically problematic. In the absence of convincing evidence that disclosure of results causes harm, I would argue that the default position should be that research participants have complete access to their own genetic data if they request it.
Absolutely right. Early on bioethicists (in the grip of an unwarranted genetic determinism themselves or fearing that the public might be) argued that such dangerous information should be kept away from people who might misunderstand it or misuse it. It turns out that ordinary people can handle even dire genetic information without freaking out.
I addressed the issue of genetic testing paternalism back in 1999 in my article, "Warning: Bioethics May Be Hazardous to Your Health."