I will also add (I am slowly reading the blog post between meetings – apologies if I have missed something) that e^abs(x) is not unambiguously e^(abs(x)) because (e^abs)(x) is a perfectly cromulent way to lexically define a function which computes first the abs of x and then the exponential. Mathematically, these coincide, but lexically these are distinct. Author intention can be determined by explicit delimiters.