Despite heading in more and more romantic directions in my thinking—from a very-analytic start—I don't find the core problem with Rand's thinking to be primacy of reason, but
sloppy (or, motivated—it can be hard to tell which) reasoning that leads to ultra-confident conclusions. A consistent pattern is you'll see a whole big edifice of reasoning out of her, but peppered about in it, and usually including right at the beginning, are these little bits that the cautious reader may notice and go "wait, that... doesn't necessarily follow" or (VERY often) "hold on, you're sneaking in a semantic argument there and it's not
per se convincing at all, on second thought" and then those issues are just
never addressed, she just keeps trucking along, so most of the individual steps might be fine but there are all these weird holes in it, so none of it really holds together.
I've even, after complaints about this were met with "you just didn't start with her fundamentals, so you didn't understand", reluctantly gone all the way to her big work on epistemology(!) and... sure enough, same.
I find similar things in basically anything hosted on the Austrian-school beloved site mises.org. IDK if this is just, like, the house style of right wing laissez faire or what.