The "so what?" sounds rhetorical and I've explained it in other posts.

Is it accurate to attribute action to a "species" or other collective image is the better question. I would say that it is less accurate to sum up specifics in a collectivized or generalized representation than it is to treat them as specific details. The problem is that cognition and representational media have complexity-limitations that result in an interest in reductionism.

Moving between the level of representation used to describe direct empirical observations and the level of generality and macro-abstraction used in certain kinds of theorizing does result in losses of accuracy and distortions. If you say, for example, that humans as a species created the idea of God, does that translate into each human individual contributed equally to its creation? Probably not, but it depends on what you mean by "contribute" "creation" and "God." If by "God" you just mean, say, knowledge of the possibility of truth, then you could say that every human individual contributes to the idea or existence of God by knowing the possibility of truth, but if you don't define God that way, then you could argue according to the specific aspect of divinity you are talking about.

Is it as accurate to talk about a species as a collective than details of specific actions of individual organisms, even if it is something that all individuals are assumed to share? No, it's still more accurate to say that "an individual labeled as Cordata has a spine" than it is to say that "the species has spines," because the logical extrapolation between conglomerate representation and individual is not left up to interpretation. The exception would be if you are talking about the species as an ideological concept apart from its application/projection on observed organisms, in which case you could say that "[the classification] Cordata contains the criteria of having a spine," but in practice people are always mixing up talking about the classification at the ideal level with talking about it as a way of representing the material entities it's supposed to apply to.

I think you could devote an entire branch of science philosophy to the slippage between these two forms of representation. The problem is that there probably already is such a branch, or maybe several, and they are all buried in libraries and other largely unvisited venues. Now, getting peer-reviewers to pay attention to this and control for it in research and publication - that would be quite a feat.

Empirically, you can observe an individual organism or entity. You cannot empirically observe multiple individuals as one. Just like you can empirically observe a single star, but you can't empirically observe a constellation. The constellation only exists as a cognitive-concept projected onto empirically observed phenomena. The words, "species" or "constellation" are themselves accurate for what they describe, but inaccurate if they are taken to describe physical entities directly instead of synthetic conceptual abstractions, which is what they are.

Is that an acceptable answer?