Excellent, excellent points. I think in this case we merely have a misplaced fairmindedness rather than 'wedge' tactics, but you're right about the effects.

Looking at scientists, too, I think you've got the right of it - they are seeking understanding, not predictions. I emphasize prediction from a kind of Bayesian-probability-theory/Popperian-falsifiability standpoint, but it's the other way around: they want understanding, and they test understanding by prediction. It's not "science explains the world in a way which lets us predict the world" - it's "science explains the world - and the explanations must let us predict the world, or they aren't explanations at all."