Thanks for the feedback. I'll go for a solid background. (ETA: Changed to timeless black to be a little easier on the eyes than some websites. Unfortunately I can't change the page's color to a light grey, will have to use some CSS. I'll consider further optimization later.)

You wording suggests you view these hypotheses as tools required to achieve some predetermined objective, rather than just as beliefs subject to observational revision.

Both and neither. I have many different epistemic practices, and I also try to switch up my epistemological approaches often. Coherentism, pragmatism, correspondence, whatever—ultimately I think the foundations of epistemology are to be found in decision theory, and any other epistemological approaches are just phenomenal shards of the fundamental nature of rationality. Hypotheses can be tools, hypotheses can be correspondences—whatever leads to intellectual fruit. "May we not forget interpretations consistent with the evidence, even at the cost of overweighting them." Similarly, may we not forget epistemologies consistent with potentially optimal decisions, even at the cost of overweighting them. We must be meta, we must be large.