> If you are an rss-reader dev then you can set up a caching layer of your own.

But are RSS reader devs willing to jump through such hoops?

I would claim that writing a (simple) RSS reader (using a programming language that provides suitable libraries) is something that would be rather easy for me, but setting up a caching layer would (because I have less knowledge about the latter topic) take a lot more research from my side concerning how to do it.

> but setting up a caching layer would (because I have less knowledge about the latter topic) take a lot more research from my side concerning how to do it.

If I was doing local development on an RSS reader, I'd just download any atom.xml file that seemed relevant, then serve it locally using php -S or some other local HTTP file-system server.

That way I can hit that file a million times without bothering any remote server. Plus, then you've made your local development environment stable even in the face of internet outages.

And if you automate the local HTTP server setup a little bit, then you can even run local integration tests.

I dunno, it seems like if "write an RSS feed reader" is an easy problem, I'd expect "serve a file from my hard drive over HTTP on localhost" should also be an easy problem.

Sure, I have done such a thing myself and it was very simple. Let's say you do http_get(rss_address). Create a function http_cached_get, that looks for a recent cached response, and if none exists delegates to http_get and saves the response. In python this is like 10 lines.