My uneducated assumption based on their docs is that it drops DOM elements or something, rather than network requests. The UI seems to be that you select things you want to be rid of, and the browser makes it so. They state that frequently-changing parts of the page, including ads, don’t get filtered, presumably because whatever they filter on is statically defined structure.
It allowed me to block the initial cookie overlay, which then allowed me to read the 'article'. Scrolling down the page triggered a popup which I could then block. Works pretty well!
I assume its blocking by origin, not behaviour? Or does that entire website just """break"""?