E.g. IQ has very little effect. You can use that assumption as cover for garbage social policy that doesn't recognize individual differences in intelligence and attempts to equalize outcomes when the distribution in ability is inherently unequal and leaves everyone less happy.
What I believe is that the interplay MAY be complex, but your argument that an unqualified assumption leads to bad outcomes works both ways.
There's no evidence for that, all the evidence so far points to linear additive models being sufficient to model the link between genes and intelligence. Here's one good overview: _Genetics and intelligence differences: five special findings_ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4270739/
What about twin studies?
It is akin to the distinction between a non-constructive proof and a constructive proof of a mathematical property of a class of objects. The former lets you work under the assumption that that property holds, but if you want to work algorithmically you need a constructive, computable proof.
Likewise, statistical inference can give you interesting signals to pursue, but it doesn't give you the models required to understand how those signals come about or how to influence them. The idea that they do is at best folly and at worst malice.
What is the consequence of assuming it? If you don't know how genes affect intelligence you don't actually have any useful theory to go by. You also don't know how to judge the effect size of genetics against other causes. These are highly non-linear, complex systems we're talking about: brains and their societies. It's far easier to use such an assumption as cover for garbage science used to oppress people who are already vulnerable than it is to turn it into a good social policy.
In short, you gain no deductive power about the world with such an assumption, just a lot of excuses for the powerful to be vicious.