Not sure what kind of cognitive capacity the dinosaurs held...

At least they probably hadn't a deceptive cognitive capacity. That is, they had few beliefs, but that few were more or less correct. I am not saying that an intelligent species is universally better at survival than a dumb species. I said that of two almost identical species with same quantity of cognition (measured by brain size or better its energy consumption or number of distinct beliefs held) which differ only in quality of cognition (i.e. correspondence of beliefs and reality), the one which is easy deluded is in a clear disadvantage.

How do you expect to recognise the former, those whose consciousness is not biologically determined?

Well, what I know about nature indicates that any physical system evolves in time respecting rigid deterministic physical laws. There is no strong evidence that living creatures form an exception. Therefore I conclude that consciousness must be physically and therefore bilogically determined. I don't expect to recognise "deterministic creatures" from "non-determinist creatures", I simply expect the latter can't exist in this world. Or maybe I even can't imagine what could it possibly mean for consciousness to be not biologically determined. From my point of view, it could mean either a very bizarre form of dualism (consciousness is separated from the material world, but by chance it reflects correctly what happens in the material world), or it could mean that the natural laws aren't entirely deterministic. But I don't call the latter possibility "free will", I call it "randomness".

Your line of thought reminds me of a class of apologetics which claim that if we have evolved by random chance, then there is no guarantee that our cognition is correct, and if our cognition is flawed, we are not able to recognise that we have evolved by random chance; therefore, holding a position that we have evolved by random chance is incoherent and God must have been involved in the process. I think this class of arguments is called "presuppositionalist", but I may be wrong.

Whatever is the name, the argument is a fallacy. That our cognition is correct is an assumption we must take, otherwise we may better not argue about anything. Although a carefully designed cognitive algorithm may have better chances to work correctly than by chance evolved cognitive algorithm, i.e. it is acceptable that p(correct|evolved)<p(correct|designed), it doesn't necessarily mean that p(evolved|correct)<p(designed|correct), which is the conclusion the presuppositionalists essentially make.

Back to your argument, you seem to implicitly hold about cognition that p(correct|deterministic)<p(correct|indeterministic), for which I can't see any reason, but even if that is valid, it isn't automatically a strong argument for indeterminism.