No doubt when we start looking at impairments and deficits that you can find plenty of situations where a high proportion of the phenotypic variance is heritable. But when you restrict the study to the top 60th to 95th percentile (which we are discussing here) I don't think you will find particularly strong relationships.