Reppert's Atheistic Argument for Violence

Victor Reppert recently posted the following blog post, which I quote in its entirety.

An atheistic argument for violence:
1) Atheism is true, and so obviously so that religious believers must be insane.
2) Insane people can do outrageous things.
3) The people who promulgate belief in God are putting other people's sanity in danger.
4) Even if we have to forcibly stop them from doing so, we can prevent them from leading other people on the road to insanity, and hence possibly outrageous actions.
5) Therefore, the use of force in the name of suppressing religion is justified.

As has been widely reported, the general public overall holds a negative opinion of the trustworthiness of atheists. Reppert is surely aware of this, so one wonders: does Reppert share that distrust? If yes, that might explain why he published a post which reinforces that distrust. If no, then why post the argument with no explanation that most atheists reject it? Better yet, why post the argument at all?  

Turning from Reppert to the argument itself, one problem with this argument is that premise (1) is false. Even if one thinks, as I do, that God does not exist and so "atheism" (in its academic, propositional sense) is true, it doesn't follow that the non-existence of God is "obvious," much less that it is so obvious that "religious believers must be insane." Moreover, even if an atheist did think the non-existence of God was "obvious," premise (1) still wouldn't follow. Instead of "insane," an atheist might instead that religious believers are simply ignorant or even stupid. Furthermore, even if an atheist did think that religious believers are "insane," premises (1)-(4) equivocate on the word "insane." To the average person, the word "insane" connotates something like "severely mentally ill," such as paranoid schizophrenia. But the word "insane" should be reserved for recognized psychiatric illnesses only, and there is no diagnosis of "theism" in the DSM-5. 

This problem also infects premise (2), which treats 'insane people' as a homogenous group. While some people with a DSM-5 diagnosis can, indeed, do "outrageous things," it doesn't follow that all or even most do. Take the risk of violent behavior. There is a significant difference in the risk profile for, say, a nonverbal person with an Autism Spectrum Disorder (ASD), and a person with paranoid schizophrenia. 

Another problem is premise (3). I suspect it is false, but I am not a psychiatrist so instead I will simply point out that (3) is not obviously true. Because (4) refers to assumptions made by (1)-(3), I think (4) is also questionable.

Finally, (5) does not follow from (1)-(4). Just because we can prevent people from leading others on the alleged road to insanity, it doesn't follow that we should. We'd need to add a normative premise, such as:

(6) If we can prevent people from leading others on the road to insanity, we should do so.

But the truth of (6) is far from obvious. It suggests that we have certainty about (i) whether a person is insane; (ii) whether insane person A will cause another B to become insane; and (iii) the magnitude of the "outrageous actions" committed by A and/or B is sufficient to justify suspending the liberty of A. But we do not have certainty about any of these things, much less all of them.

In short, the argument for violence fails.