WE have to remember that these twin studies all share the same methodology - a cinfusion between hypothesis testing and parameter estimation(just because a parement is called G doesn't mean that it is genetic; these studies assume the very result they are trying to prove);  a lack of experiemntal controls (opposite sex v. same sex DZ twins for example) and a lack of stanrd errors (more than half the time the hetirbaility estimates are not statitiscally signficant).
But as we have learned over the last 30 years we dealing with a form of religious belief called by Dick Lewontin in another context,  Darwinian fundamentalism. Such beliefs are not moved by evidence and argument.