ELI5: Prior to Safari's new content-blocking API, browser extensions would detect if a resource was loaded, and then either:
modify the css so it disappears,
inject scripts to modify a website's behavior,
or prevent scripts from executing
This method requires a lot of invasive and complex javascript running on your page to dynamically modify the behavior of a website. Many of these extensions (in the beginning) used lots of additional resources by including libraries and javascript that would slow down load time and increase memory consumption. This problem became worse as companies fought back against these ad-blockers by making their ad's more difficult to block, so more code was needed.. and this caused a split to occur:
1) Advertisers began to 'work with' adblockers (secret whitelists), and advertisers began building their own blockers with catchy logos and good marketing strategies (ghostery is the best example).
2) Indie-dev, open-source, adblockers (like ublock(origin)) that gave users complete control, focused on decreasing resources and memory consumption, and relied on 3rd party lists updated weekly, with a flexible syntax that let them combine multiple ecosystems.
Some aspects of #1 aren't bad. Adblockers with whitelists forced advertisers to fundamentally change how they do ads to get on the whitelist (unless you're super rich and powerful). But it also represents a compromise, where there can be some ad revenue, and it's not the worst thing in the world. Ghostery, on the other hand, is fascinating to me. They are openly working with advertiser, it is no secret, it is explicit. But, they're giving advertisers information about users who are blocking ads that benefit the advertisers. They claim it's to make 'ads better', but, the reality is the information ghostery has access to is far more valuable than ad revenue. Ghostery doesn't lie, at all, but they have a powerful adblocker, because they have the consent of the advertisers and they're giving them something valuable. It may be harmless, semi-anonymous, non-creepy, just people wanting to make money, and supposedly you can opt out. But, it's completely based on trust, and there's no guarantee that the practices will remain consistent, and we have no idea what level of activity is being tracked, so.. I'll be optimistic and assume Ghostery is good and a compromise.
Option #2 is the death of advertising. It's their worst nightmare. It fundamentally gives us leverage over them, but we might inadvertently just be screwing over the people in middle, the ones who build the website we enjoy using. So, I support it, it's obviously the better choice for consumers, and it will basically force websites to make money via other means (such as donations).
So, Safari content blocking is just a JSON file - a list of websites with some elements and regex, that's all. It cannot be used to invade privacy, it literally can only block content. Safari decides what can loaded or shown on a page in the networking / rendering layer of the pipeline (that's my best guess). So it's blocking the content one layer before a normal extension would, and it's doing it without custom code by the developer. That's why a page load time will decrease from 7 seconds to 1 second, because the resources aren't even being loaded. With a normal blocker, it can stop the execution, but in general, the website get's a lot closer to loading the resource, and in most cases it's loaded and not displayed. Safari is like.. 'I ain't even gonna try to load that **** and I'll cache the page without that resource'.
But.. if a website is expecting a javascript event to occur which signals an ad was observed, loaded, or clicked, that's not programmable in a JSON file. So, the website will break. And that's what we see occurring in many situations with the new safari blockers. So, although they're fundamentally more powerful, it's trivially easy for a website to 'protest' a content blocker by simply not loading their site if an was blocked.