A theory can be useful without being the final description of what exists.

Usefulness and truth are different tests. A model can predict well, guide good decisions, and hold together beautifully while still being a simplification that a later account will replace, and it loses none of its practical value for that. Demanding finality from ideas that only need to work is a way of misunderstanding what ideas are for.

Hold your own frameworks the same way, firmly enough to act on and loosely enough to revise. When something you believe is challenged, the question is not whether it is ultimately true but whether it is still doing better than the alternatives. That posture makes you both more decisive and much easier to correct.