Scientific papers usually define what they are studying.

When I say concrete meaning I mean that sentience is an abstract concept of which we can observe evidence of, but we cannot define clearly what it is. In the report you mentioned, you will see that they give 8 criteria for scientific evidence of sentience, i.e. these do not define what sentience is, but they are criteria that we presume sentient beings should satisfy. They even require several pages to explain the complications of how to define sentience and how to observe it.

I do admit that the extent of study on sentience of animals is greater than I initially thought, and I can see that one might have reasonably sufficient knowledge to judge, with some certainty, which life organism might be sentient (under definitions such as the one used in the report). But it seems to me nearly all animals fall under this umbrella of “some level of sentience”, I found this paper highlighting that many insects seem to have cognitive abilities, and might be capable of feeling harm. So to what extent must this go, can you not swat a mosquito in fear of its suffering?

If I throw a rock to the ground, it doesn’t make sense to say I harmed the rock, because a rock can’t experience being harmed

But a rock is not alive, there is no evolutionary force driving its interest, as with all other living organisms. A sea cucumber has no proper nervous system (as I understand from a quick search), and thus could not “feel” pain. Yet, if you cut one in half, I would say that you have harmed it. But this is really just discussing the semantics of the word “harm”, the real point is that you are doing something to the organism that goes against its natural interests.

If we do not follow speciest dogma, we might as well eat other humans. Indeed meat eaters don’t really have good reasons to exclude human meat.

Yes they do, speciesism. A quite natural reason.