Thanks for your version of the Consequence Argument.
Let me try to distinguish what I call the Cogito model of free will from yours/van Inwagen’s and Kane’s accounts.
I believe my model will appeal to many determinist/compatibilists because the results are only different where the randomness allows us to come up with an unpredictable and creative new idea.
PvI pointed out that a key discriminator is to see how free will models handle “exactly the same circumstances,” something Robert Kane calls “dual rational control.”
Given that A was the rational choice, how can one defend doing B under exactly the same circumstances?
PvI’s description, for example, assumes that chance is the direct cause of action.
van Inwagen says:
“Now let us suppose that God a thousand times caused the universe to revert to exactly the state it was in at t1 (and let us suppose that we are somehow suitably placed, metaphysically speaking, to observe the whole sequence of “replays”). What would have happened? What should we expect to observe? Well, again, we can’t say what would have happened, but we can say what would probably have happened: sometimes Alice would have lied and sometimes she would have told the truth. As the number of “replays” increases, we observers shall — almost certainly — observe the ratio of the outcome “truth” to the outcome “lie” settling down to, converging on, some value. We may, for example, observe that, after a fairly large number of replays, Alice lies in thirty percent of the replays and tells the truth in seventy percent of them—and that the figures ‘thirty percent’ and ’seventy percent’ become more and more accurate as the number of replays increases. But let us imagine the simplest case: we observe that Alice tells the truth in about half the replays and lies in about half the replays. If, after one hundred replays, Alice has told the truth fifty-three times and has lied forty-eight times, we’d begin strongly to suspect that the figures after a thousand replays would look something like this: Alice has told the truth four hundred and ninety-three times and has lied five hundred and eight times. Let us suppose that these are indeed the figures after a thousand  replays. Is it not true that as we watch the number of replays increase we shall become convinced that what will happen in the next replay is a matter of chance.”
(”Free Will Remains a Mystery,” in Philosophical Perspectives, vol. 14, 2000, p.14)
I think this agrees with your points 4), 5) and 6), no choice about – or control over – the decision, thus no FW or MR.
For more on van Inwagen, please see http://www.informationphilosopher.com/solutions/philosophers/vaninwagen/
Robert Kane is well aware of the problem that chance reduces moral responsibility, especially in his sense of Ultimate Responsibility (UR).
In order to keep some randomness, Kane says perhaps only some small percentage of decisions will be random, thus breaking the deterministic causal chain, but keeping most decisions predictable. Laura Ekstrom and others follow Kane with some indeterminism in the decision.
Let’s say randomness enters Kane’s decisions only ten percent of the time. The other ninety percent of the time, determinism is at work. In those cases, presumably Alice tells the truth. Then Alice’s thirty percent of random lies in van Inwagen’s first example would become a mere three percent.
But this in no way explains moral responsibility for those few cases.
For more on Kane, please see
Compare the Information Philosophy model, which agrees with compatibilism/determinism except in cases where something genuinely new and valuable emerges as a consequence of randomness.
In our two-stage model, we have first “free” – random possibilities, then “will” – adequately determined selection of the best option.
Alice’s random generation of alternative possibilities will include seventy percent of options that are truth-telling, and thirty percent lies.
Alice’s will evaluates these possibilities based on her character, values, and current desires.
In the Information Philosopher’s Cogito model, she will almost certainly tell the truth. So it predicts almost the same outcome as a compatibilist/determinist model.
It is not identical, however, merely adequate determinism.
It is possible that among the genuinely new alternative possibilities generated, there will be one that determinism could not have produced.
It may be that Alice will find that option consistent with her character, values, desires, and the current situation she is in. That might include a pragmatic lie, to stay with van Inwagen’s example.
In a more positive example, it may include a creative new idea that information-preserving determinism could not produce.
Alice’s thinking might bring new information into the universe. And she can legitimately accept praise (or blame) for that new action or thought that originates with her.
For more on my model, please see http://www.informationphilosopher.com/freedom/cogito/
To summarize the results:
……………. / Van Inwagen / Kane / Doyle / Compatibilism /
Alice tells truth /…. 70% /…. 97% /…. 100% /…. 100% /
Alice lies……. /…. 30% /…. 3% /…. 0%* /…. 0% /
* (unless a good reason emerges)
Let’s consider the Moral Luck criticism of actions that have a random component in their source.
Alfred Mele would perhaps object that the alternative possibilities depend on luck, and that this compromises moral responsibility.
Mele may be right with respect to moral responsibility.
In any case, free will and creativity may very well depend on fortuitous circumstances, having the new idea “coming to mind” at the right time, as he says.
The universe we live in includes chance and therefore luck, including moral luck, is very real.