That's not true, for example metacognitive ability studies have shown environment plays the dominant role. Twin studies on trust provide the same, in which genetic component while large at 33% certainly doesn't indicate what you're stating that it's only "extreme cases". Even in studies re-assessing conventional twin studies and educational attainment, the conclusion was that while some is genetic (sometimes even a large portion) the correlations between a mother and father's educational attainment points to environment playing a large role (unless you have the belief that the mother and father are siblings I suppose).

You'll be extremely hard pressed to find researchers conducting these twin studies who minimize the role of either genetic or environmental impact on certain aspects in the way you did.