It requires a pretty die-hard idealist stance to claim that the mixed state remains mixed until some consciousness comes into contact with it. I tend to prefer a "softer" role of consciousness-- the concept of a hypothetical consciousness suffices. That handles the "tree falling in the woods" issue. The point is, most scientists are too realist to want to imagine that a consciousness is actually required to collapse a wave function, but one does need to introduce a hypothetical consciousness into the problem to collapse the mixed state. The way that looks is, we say the mixed state is "actually" one result or the other, because we can imagine introducing a hypothetical consciousness/intelligence/analyzing agent into the situation without disrupting the system (because it is already highly macroscopic) to adjudicate which outcome occured. It doesn't matter if the consciousness is there, what matters is that our consciousness can imagine that one.
The collapse then occurs via a kind of "mini me" mindset, much like the way "which way" information collapses a two-slit experiment whether or not any consciousness is actually there to interpret that information. Note that in this brand of interpretation, the consciousness is still required to give meaning to the experiment, it is just not required to be on hand in the actual apparatus. This approach allows for realism, which most scientists view as a convenient philosophical stance whenever possible.
This doesn't present any problem. The idealist simply holds that the universe is in a kind of unactualized state until the consciousnesses come along to collapse it into a definite state. But the realist finds that too radical, yet can still easily find a role for consciousness in the same "mini me" fashion I mentioned above-- we simply say that because we can imagine a consciousness being present without altering the reality, it doesn't matter if the consciousness was really there or not. Note this is what you cannot do inside an atom, say-- there is no way to imagine a consciousness present inside an atom without changing the atom, the atom has not been rendered into a macroscopic mixed state that the (hypothetical) consciousness could collapse.
Now I know what some might say-- they'll say that a hypothetical entity cannot actually "do" anything in a physical situation. But that is because they have already adopted a fairly radical philosophical stance that reality can actually be parsed into "doers" that cause things to happen, and "onlookers" that have no effect. That approach has worked for us for so long that we forgot to notice how radical a stance it it, but it is in quantum mechanics that this stance falls all apart. Instead, in quantum mechanics we find that even the onlooker is responsible for the information they are using in their physics-- their fingerprints are all over the way they are choosing to think about any situation, the coherences they are choosing to track and those they are choosing to average over and ignore. Quantum mechanics is simply not possible without making choices like that.
The analogy doesn't hold-- we don't need Santa to do science. We do need consciousnesses/intelligences/analyzing agents (again I make no effort to parse any distinctions there), they are demonstrable and inextricable elements of doing physics. We just didn't need to think about them before quantum mechanics, because we never faced the question, "where does the definite outcome come from?" That is exactly the question that forces us to adopt an interpretation of quantum mechanics where we did not have to adopt an interpretation of classical mechanics, because we never included any indefinite outcomes in our theories before.