My understanding of his reasoning is that he does credit Musk with knowing things about thermal expansion, and it's precisely that combined with the hand-wavy t

My understanding of his reasoning is that he does credit Musk with knowing things about thermal expansion, and it's precisely that combined with the hand-wavy treatment of it in the proposal that makes him believe that the whole thing is likely to be snake oil. He's not really trying to have a reasonable argument, he's describing his thought processes around whether this is a project that is worth him getting involved in. There are a lot of bad ideas out there, and geniuses (if that is what Musk is) are no stranger to them (c.f. Einstein's wing design). It's perfectly reasonable when faced with a huge number of proposals that take a significant amount of effort each to make an initial determination based on surface details. Sure you'll have some false negatives, but you'll have a bit more time to address the ones that really have a chance of going somewhere.