Again, the timescale issue. If the culture doesn’t contribute to reproductive success, it is waning. If it becomes detrimental to reproductive success it can doom the population. Over a long enough timescale, a culture cannot remain static while not contributing to reproductive success in a dynamic fitness landscape.Reproductive success is indeed necessary for the continuation of the population, but that doesn't say that a population that has culture cannot thrive if the culture itself doesn't contribute to reproductive success.
/unsupported opinion
Good thought. There’s usually an upside/downside to any adaptation. Even if the adaptation itself is uniformly (reproductively) beneficial, it almost always comes at a cost.Or it always was maladaptive. This may be more common than we intuitively think, though this also depends on the granularity of the analysis. Tuning down the immune system during pregnancy is arguably adaptive as it allows the fetus to escape being eliminated - but there wasn't a point in our evolutionary history when pregnant women being more susceptible to infectious diseases than the population average was adaptive.
It's going to be a fuzzy border. We're talking biology after all.
It (distinguishing the learned from the instinctive) may be a totally intractable problem, as I suspect.
A complete answer might not even contribute to our overall understanding of behaviors, even if couched in terms of population genetics.