They say this can't be done with something on the scale of a human, but I don't understand why based on Wiki articles on the Schrodinger's Cat and Wigner's Friend thought experiments. This question isn't that related to the article but, since the result of every experiment is determined by a perception, how do we know that quantum superposition is a property of the atoms in such an experiment as opposed to being a property of our perceptions of the experiment?

Here's a modified version of the Cat thought experiment that perhaps accentuates my confusion:
- In the original thought experiment, if the device is triggered then the poison is released and the cat is killed. For humanity's sake, let's say the device instead triggers a wooden block to be dropped by a robotic arm. Say that we modify the device so that the block is dropped at a specific time (instead of at a random time) if the random quantum process occurred within some specified time period T. In this experiment, if we hear the block hit the ground at time T then we have effectively observed the outcome of the experiment and the superposition has collapsed.
- Now, say we mount a second device in the box which is not connected to a the random quantum system, but instead drops a wooden tetrahedron at time T if the first device is not triggered. So, at this point, I see that the second device is also coupled to the random system despite my saying that it isn't, but the point of the second device is to produce a decoy sound.
- So now, if at time T the listener hears a "thud" sound, he/she can no longer conclude the outcome of the experiment with respect to the wooden block. Does this mean that the initial device is still in quantum superposition? In other words, is superposition collapse caused by interaction between the listener and the experiment ("observation" in the conventional sense) even if that interaction doesn't result in actual knowledge of the outcome, or does it only occur when that interaction is such that the experimenter is certain of the outcome? If the latter is true, then superposition collapse results only from observations that distinguish the possible outcomes.

Follow-up questions:
1. Does the superposition still collapse if the listener thinks he/she knows the outcome (e.g. if they didn't know the decoy device had been added).
2. Does the superposition still collapse if the listener has enough information to distinguish the possible outcomes but simply fails to put the evidence together correctly (since accidentally observing the wrong outcome does not constitute legitimate certainty) or is simply too apathetic to do so?

In summary my point is this: Since observations that quantum effects do indeed exist can always be traced back to some perception verifying the quantum effects, how can we be sure that quantum effects are not a property of perceptions themselves when one is perceiving under particular circumstances? I think the main flaw in my thinking is that I fail to acknowledge the distinction between quantum uncertainty and classical uncertainty. Does Schrodinger's original experiment even allow for that distinction? Someone asks that same question here. Apparently the answer is "No" (for the dead-alive superposition, at least).