I kinda switched off before the flame-bait portion even started.

The first few paragraphs of text appear to be confusing Lisp/Scheme with 'functional programming' as a paradigm, and making an argument about one using the other. There would be nothing easier about "modifying the way the compiler generates code" in OCaml than in C or anything else (assuming a well-written compiler in each case).