Most times, I am making HTTP requests with TCP clients, not a browser. If I am using an HTTP client, I use tnftp with custom modifications, busybox wget, or some other small client, not cURL.
When I retrieve HTML/JSON/CSV, I use small filters I wrote myself to extract what I want and present it in a readable text-only format and/or insert it into a SQLite3 database. I like the sqlite3 text-only output formats.
Hitting a limit on the number of news articles I can read is not something I am personally familiar with, and I use the same news websites as the people who complain about this, hence I believe it is associated with the browsers they are using, not the websites.
I operate in textmode, no graphics. There are few benefits I get from the use of HTML without useful hyperlinks.
In such cases, using a browser is unecessary, overkill. I use a pager instead.
Or use a client that does not implement cookies, localStorage or Javascript, like the original www browsers. I have been using these for decades.
In the relatively rare case I have to use a popular, graphical browser distributed by an entity that collects users' data to support online ad services, on a network I do not control, I use Javascript site settings and/or UBlock Origin. But that is far more complicated, more resource intensive and slower compared to using a simpler client and a localhost-bound forward proxy. Plus I am at the mercy of third parties: 1. the advertising-supported company distributing the browser and 2. the "browser extension" developer.