You are mostly right. There is some additional value provided:
With known feed urls e.g. https://www.meetup.com/%s/events/rss/ we can save the network call to meetup.com and produce the feed url immediately. This is also an argument for making a browser addon or at least some client code that can produce it.
Discovered feed urls are cached which saves the direct network call to a possibly slow site.
The service produces third-party feeds for websites that don't have it e.g. https://twitter.com/elonmusk => https://nitter.net/elonmusk/rss
