CFLarsen said:
Please tell me you are kidding. Did you not understand the analogy?
Let me try with a coin example. The RPKP experiment would go something like this: I am going to flip a coin 1000 times.
While I am flipping it, please try to make it come up heads more often than tails. At the end, we will count how many excessive heads there were, and calculate the probability of that happening due to chance.
Your suggested experiment might go something like this: I am going to flip a two-headed coin 1000 times. Please make it come up tails JUST ONCE.
Or maybe this: I am going to make a fair coin come up tails EVERY time (miraculously!). Please make it come up heads JUST ONCE (what?).
Suppose I made the claim that, by blowing on the coin mid-flight, I could cause heads to come up more often than should happen with no external influence (in the first experiment). Would you design experiments 2 and 3 to test this hypothesis?
Do you see why these experiments are fundamentally different?
Such a claim can ONLY be tested probabilistically. After the probability of it happening by chance has exceeded some constant (1 in a million? Billion? Trillion?), it would start to become foolhardy to claim that nothing is going on. How do your proposed experiments reveal anything at all?
EDIT: I will give you the benefit of the doubt, and assume it is the quantum mechanics portion that isn't clear to you. The stream of qubits being generated by the RNG can be seen as a stream of coins flipping in the air. Not until a measurement is made on a qubit does it "land." There is nothing being "changed," so no 0s can be "changed" to 1s. The qubits merely haven't "landed" yet.
(caveat: Whether or not the coherency of the qubit can be maintained in a classical stream is a quite separate discussion.)