I think you need to distinguish between different kinds of "using a theory." Investigating something you don't believe in is not the same thing as using (perhaps with reservation) a model you believe to be imperfect, and that in turn is different from claiming as true things that you believe to be totally false.
In general, a scientist will never "claim as true" anything. They will make claims of the following nature:
Theoretical: Given assumptions A, we derive conclusion Y.
Experimental: We observe phenomena Y. This is consistent with A, but inconsistent with A2. Thus, assumption A2 must be false.
The fact that you don't believe theory A doesn't make your experiment refuting A2 invalid. And the fact that you don't believe A doesn't mean that it's incorrect to make conclusions about what would happen if they were true.
And in particular, none of this has any bearing on holding beliefs which are not even wrong. You might believe in Copenhagen QM, I might be a Bohmian, but that doesn't make either of us frauds when we do calculations involving the Schrodinger equation. That's the thing - science isn't religion. The correctness of a theory or the usefulness of an experiment is not dependent on the beliefs of the person performing it.
"and that in turn is different from claiming as true things that you believe to be totally false."
Dawkins approach would lead to more people claiming as true things they believe to be false, so they can get the jobs they want. Kind of like how people living under totalitarian regimes pretend to really like government policies and government officials, in order to get or keep a favorable position within the system.
If it seemed like I was arguing for Dawkins' position, that wasn't my intention. It just seemed like yummyfajitas was lumping a lot of very different things under a vague umbrella term, which I think does a disservice to the argument he was making.
