To refute your last point, the article explicitly mentions that that Twitter has functional RSS support, which was deliberately removed from the user interface and metadata, so it's not a matter of development or maintenance cost.
"The demise of RSS"
"The downfall of RSS"
Or if you want to get superdramatic: "The silent acquiescence of RSS to an untimely demise"
It’s not a war. Not even using extremely liberal (and stupid) definitions of war.
Mechanical fish is right, RSS is unattractive to people, that is the problem. People don't hate it, it isn't a war.
<link rel="alternate" type="application/rss+xml" href="...">
pointing to a URL that already exists and works just fine doesn't cost anything. If the primary reason had been maintenance cost, they would have removed the feature itself.
The next step, in many cases, is to send out sorrow-filled notices to the feature's users saying, in essence, that you're end-of-lifing the feature as of date X. Though often the bad news is artfully disguised. (e.g. "Look at all the glorious new stuff that we're rolling out to replace that tired old feature that will now be deprecated and slated for removal.")
Only in step three do you just shut it off. You might even shut it off gradually, as when Google shut off their old design for Gmail, a process which is still going on and appears to be taking literally years.
How much time elapses between the three steps depends on the details.