Why isn't it exactly the same as if the doc had the includes filled out and then processed semantically, much like C?

All <<<include whatever.html>>> could be replaced with contents first. Then it's just a page of html as now.