They call for further study at the end, acknowledging that their data was all /in vivo/ in a lab, and not direct data in humans. Thus, while they claim to have demonstrated a mechanism, they make no claims that it works the same way in people.
Tangent: most papers seem to do that… which confuses me. It feels like it’s basically saying to discard everything the paper talked about because to be actually sure you’d need to investigate more than they did.
Is it a pro forma formula or does it really mean what I understand it to?
Generally, answering questions will raise new ones. It is very unlikely for a single paper to be the end of a line of research. Sometimes the new questions are quite fundamental (like here), sometimes the new questions are details, and sometimes the further research is "try this same method on this other known big problem".
Besides the inherent fractal nature of how science works, there is also an incentive. Being a scientist in a solved field is not very smart. Or even more cynically the 'further work' section says 'please pay me to do these things next'.
But I think the cynical view is not as important here. Research tends to be strictly limited in time, and for good reason. During the research you will probably find new directions, and also get insight into what old questions are interesting, or what new methods matter. All of these are great cause to include a further work section.
