Sorry, I left out the part where most RSS fetchers are not hosted by the user. Of course it is self-hostable, but that’s by far the less common use case.

Images and CSS aren’t natively a part of RSS, though (and in fact I don’t think I’ve ever seen an RSS feed or reader that tries to do any CSS rendering at all). Assuming you have a third party downloading your RSS XML, all of the tracking capabilities are outside of the RSS spec itself, and dependent on you clicking on a link or something after you get the RSS feed.