I read The Grand Design. I'm long acquainted with much of the history and physics therein, albeit at a conceptual not mathematical level. However, I was fascinated by the description of the entire universe as a Feynman path. I can't make any knowledgeable comments on that or the M-theory stuff, of course. I couldn't help wondering if the sections on renormalization and "negative energy" would've been easier to understand with the careful and hand-held inclusion of some undergraduate-level math. That's a hard balance to strike, though. Maybe I'll try some cross-referencing with the tome that's "heavy" in several senses of the word, The Road To Reality by Penrose. I doubt the two books share the same general opinions.
Since I'm monotonous, I'm obligated to compare the book's "model dependent realism" with my interpretation of philosophical Pragmatism. I noticed many similarities. In model dependent realism, humans perceive reality through the lens of a model. In Pragmatism, humans perceive reality through the lens of subjective elements like desire, focus, analysis, synthesis, theory-building, etc. In model dependent realism, humans select models for the sake of "convenience". In Pragmatism, the convenience of thoughts about reality is explicitly tied to how well the thoughts "work" for purposes. In model dependent realism, humans replace models as they compare the accuracy by experiment. In Pragmatism, humans adjust their knowledge of truth as they actively determine which individual truths are confirmed "in practice". Most infamously, in model dependent realism, an ultimate universal model of reality might simply be impossible, except as a quilted combination of an array of limited models. Just as infamously, in Pragmatism, truth isn't a standalone all-encompassing entity, except as an evolving collection of ideas whose two coauthors are the human and their whole environment.