H/T Ophelia Benson and her commenters.
This topic is, of course, never ending. It's been brought up this time over at the Huffington Post, by Alan I. Leshner, CEO of the American Association for the Advancement of Science, and Executive Publisher of Science, .
As usual with accommodationists, there is no real understanding of what non-accommodationists actually say, which has its nuances. What we say isn't just "religion and science are incompatible", which is ambiguous, and could mean various things that are false. We do say that, but we go on to gloss what we mean by it.
In my case, what I say is something like this: they are incompatible in a sense. Accordingly, it is misleading to state simply “science and religion are compatible” as if there's no problem. If you say that, you'd better gloss it, and you'd better acknowledge that, in the sense that actually matters to traditionally religious people, they may not be compatible, and that there's thus a big problem. When I was a religious person, I didn't care whether it was psychologically possible for some or even many people to be both (a) scientists and (b)religious. I cared about the consistency between (1) the truth claims of the sort of the religion I subscribed to and (2) the more robust truth claims of science, and inferences that could be reached from these together with other fairly plausible premises.
The position as I see it is something like this: viewed historically, religion needs to thin out its epistemic content, or to introduce notions of the capricious way supernatural beings act, or to adopt intellectually unacceptable ad hoc tactics of various kinds, in order to maintain a formal compatibility with the scientific picture of the world; the advance of science pushes God into smaller gaps; and some religious views are plainly inconsistent with robust scientific findings. All this reflects a general mismatch between the scientific approach to the world and the religious approach, which follows from (1) the fact that they use different methods for discovering the truth and (2) the methods of science do not, historically and contingently, reach the same conclusions as previously reached by religion. It turns out that religion needs to adapt constantly, thinning out its original truth-claims or making various ad hoc manoeuvres, or it find itself plainly contradicted by science.
All of this then feeds into arguments that the religions of the world are probably false across the board. The evidence is that they use unreliable means of looking for knowledge - but why, if they have access to gods, angels, etc.? Meanwhile, various specific religions are already falsified to the extent they are plainly or less plainly inconsistent with robust elements of the scientific picture of the world.
As far as I know, non-accommodationists like me, Jerry Coyne, and Ophelia Benson all see it much the same way. We are not clones of each other, and our views may have minor differences. Jerry and Ophelia might not sign on to every word in this post. But we are all prepared to spell out something like the above story – and we have done so on various occasions. We don’t claim that no religious view could ever be logically consistent with what is known by science at a particular point in history - that's palpably false because some religious views, such as an extreme Deism, are so thinned out as to be unfalsifiable - but we do object to simplistic, misleading claims that “science and religion are compatible”.
Nothing about the non-accommodationist accounts is falsified by telling us, over and over, that a lot of scientists are religious. That's really about psychological compatibility. Note, though, that the evidence suggests that a far higher proportion of scientists are not religious, compared with the general population.
Moreover, those who are religious tend to have a very thinned out view - in this survey data only 3.7 per cent of scientists found most truth in a particular religion! By contrast, 73.5 per cent thought there was basic truth in many religions, but one must wonder what those basic truths are, since 62.2 per cent either just plain do not believe in God or say they don't know and there's no way of finding out. Those figures must overlap heavily, so it looks as if the sort of truth that many scientists find in religion is most likely some kind of generic morality or the like. (Only 9.7 per cent believe in God without any doubts and only about 23 per cent attend church more than 1-5 times per year.)
More research needs to be done, but the data we have is totally consistent with the non-accommodationist idea that science tends to push people either away from religion entirely or into some sort of "thin" religion with little of the traditional content. That is not going to comfort religious people who are suspicious of science, and nor will it comfort those accomodationists who want to paint the picture that there's just no problem. For what it's worth, the data we have favour the non-accommodationist position, once the latter is understood - and not represented by a straw man version.
Frankly, I think the better evidence is what you get when you simply place the claims of various religions side by side with the more robust findings of science. Given what we now know, do the religious claims seem plausible or not?