Regardless, I take issue with your contention and seemingly fundamental assumption that the article's original title is somehow better and will win out against alternative titles. The entire reason people get angry about this is that the alternative titles are often much more useful: they are more precise, and they are even sometimes less sensationalized.
Sometimes titles just need to be different when the material is being viewed in a different context. I've actually been chewed out before for "editorializing the title" when I wrote the original article in a case where my original didn't even have a title because the medium (Google+) doesn't use them. In other cases, the in-context title might need to be very short, but on Hacker News it should be longer.
The places you thereby see users actually complaining about this policy are not the places where the submitted title sucked. To demonstrate just how dismissive it is to claim that that is the situation we are discussing: it's like trying to end a debate on whether airport security should strip search anyone who talks about security issues because clearly everyone who does so is a dangerous terrorist.
In fact, even people (such as myself) who complain about title changes made on Hacker News often call for title changes, as to many of us the problem is "inaccuracy", not "originality": if the title says something that doesn't seem to be true (as in the case just yesterday regarding the Tesla S's usage of rare earth metals) it isn't the person submitting the link that is "editorializing", it is the original author.
Another interesting circumstance was one of the original cases from about a year ago that made this turn into a big public argument: the visualization on the New York Times about some non-hacker aspect of politics. The visualization was insanely impressive, and someone wanted to submit the visualization. The title was clear, factual, and even quite concise: it just labeled the page as being a visualization.
In this case, there were tons of interesting comments about the visualization, and people were really excited to see it: the link was clearly a success, it was voted to the front page, and not a single person was bothered that it was there... until some moderator changed the title to the original title of the visualization, which was about some political issue, and now the comments changed in tone.
Now, people were arguing about how this wasn't worthy of being on Hacker News: that political issues should be discussed elsewhere; then a fight breaks out about the underlying politics. People are infuriated that somehow this link ended up on the front page with an insane number of votes, and no one has any context due to the policy of changing the titles of posts, even after many hours of discussion in hundreds of comments, without any historical notice that it happened.
This is the kind of situation that I think it is reasonable to be bothered by; sure, there are some people who believe that the titles should never be changed under any circumstance, and to the extent to which they exist it is convenient to lump them together with everyone else because "well, people editorialize and there's linkbait so the titles should be changed QED"; but that's not the argument that I've actually seen about title changes: people just want more nuance and more transparency.