If we can suspect X given Y, but we shouldn't build models on top of the assumption of X, then what are we supposed to do with Y? To me it seems like you're ar

If we can suspect X given Y, but we shouldn't build models on top of the assumption of X, then what are we supposed to do with Y? To me it seems like you're arguing that it was a bad idea to build on the assumption of Newton's theory of gravity because eventually it would be replaced by Einstein's theories of relativity. Which is obviously not sensible, since Einstein's theories were in part the result of trying to explain inaccurate predictions made by building on Newton's theory.