Earlier quoted context omitted.
I'm just an amateur physicist... but since you're a pro, I'm sure you don't really mean it when you say Einstein 'used a theory he didn't believe was correct (quantum mechanics) to derive conclusions he didn't believe were true'. I mean, that could be misunderstood quite easily as the pop-physics version of history, in which the aging old Einstein just can't hack it, staying stuck in the old ways of thinking. In real…
I mean, that could be misunderstood quite easily as the pop-physics version of history, in which the aging old Einstein just can't hack it, staying stuck in the old ways of thinking. This is totally the opposite of what I meant to convey. Einstein/Podolsky/Rosen were doing great work. They helped us understand an important implication of quantum mechanics - that if our configuration-space based theories are correct,…
Now, when you say he derived conclusions he didn't believe were true, that's kind of wrong, since all he derived was a dilemma (completeness or separation, not both (but possibly neither)). He took this to show that if we maintain the separation principle, then QM needs to be viewed as a statistical account of objects with properties it can't describe (but he never took that to mean hidden variable theories like Bohm's were the answer).
So, my point is, he believed in the correctness of QM (just not in the completeness). He believed in QM being true - just not that is was the whole truth. He believed in the correctness of the EPR derivation and in the correctness of the dilemma at the end. When you say he used a theory he didn't believe was correct, I think that's wrong - he just didn't believe it was the finished article. And when you say he derived conclusions he didn't believe were true, I think that's wrong too, since he derived a dilemma that he very much believed in. In the end he chose the wrong side, but that's not relevant here.
You say Einstein believed this was a reductio ad absurdum - but that's flat out wrong I think. EPR relates to the completeness of the formalism (if the separation principle is taken to be true), not the correctness.