"In that case I’d expect Twitter to argue that the feature they want to remove for engineering reasons is filtering out some of the tweets you see based on whether you are a follower of the person the message is directed to not the other way around. "
That would be easier for the twitter engineers, but thats not what 97% of twitter users want. The author doesn't acknowledge this.
That 97% only indicates that most people don't change the defaults; it says nothing about what Twitter users want.
In this case it's particularly misleading, since we're talking about an option to turn off silent data loss. What would make someone track down a setting related to a problem they don't know they have? A more honest representation would be: "A small number of our users were aware that we filter @replies by default, some were happy about it and a few (3% of total users) changed their settings to remove the filtering."
That would be easier for the twitter engineers, but thats not what 97% of twitter users want. The author doesn't acknowledge this.