The easiest way to identify most of them, is a means to track every address and see where the loading process stalls at and if not essential, add into the deny / host file. This usually works but some sites will have a hissy fit when they don't get confirmation from some minor resource hasn't had a successful scrape and then refuses to load at all, or loads and blanks out the page. A few are pretty crafty like that - it doesn't solve everything and sometimes there is a price. I have banned one AI driven script a couple of years ago, that immediately afterwards it was like flicking a switch and feeling like a fifth of the net wasn't available any more - I can live with that though.
[1] tehparadox RIP 2017, of which in years afterwards many new versions popped up but were not viewed as valid as far as a couple of the old staff I knew from tehparadox