I agree, but would modify it.
The hypothesis doesn’t have to answer the questions, it has to make testable, empirical predictions about the answers to one or more of those questions. One can easily formulate ID hypotheses that do so, but those must have the potential for falsification, so no ID proponents will go there.
Gilbert is omitting the essence of science.