So you think assuming a less likely position makes people better scientists? I'd like to see a calculation that a system of optimization that proceeds with that assumption actually does converge...
Of course one should be open to other possibilities. However, ones openness to a possibility should be related to the likelihood of that possibility. If that's not the case, you are disconnected from reality.
>So you think assuming a less likely position makes people better scientists?
You're twisting my words. I think believing that certain assumptions which are required to support your work are false makes you a better scientist. The specific person Dawkins maligns hasn't taken any positions in their work that align with his personal beliefs. But I believe that his work is better for it.
Really though, this is more a question of diversity of opinion than of one belief system being intrinsically superior to another for the doing of science. We are all disconnected from reality in one way or another. If all of our scientists are disconnected from reality in a similar manner, it's going to drastically stunt their thinking. And this is equally true if all scientists have the same sort of atheist-agnostic belief that science is mostly right in its models or if all scientists are young-Earth creationists. (At least in a hypothetical world where all scientists are YEC and simultaneously produce rigorous work that makes YEC appear almost certainly false.)
So long as their work is rigorous and aligns with data, their personal beliefs merely serve to give them novel ideas Dawkins could never think of.
Of course one should be open to other possibilities. However, ones openness to a possibility should be related to the likelihood of that possibility. If that's not the case, you are disconnected from reality.