So why not have, for _every_ story, a "History" link, or even a _feed_ of latest changes, so that readers who re-visit a story can quickly see the atomic updates? For my own reading habits, this would cement the NYT has a place to go to for developing stories...the transparency/accuracy bit are just icing on the cake.
This issue of having an in-house Newsdiffs was brought up by the public editor back when Jill Abramson was still editor:
http://www.nytimes.com/2011/06/26/opinion/sunday/26pubed.htm...
Here's what she said:
> Right now, tracking changes is not a priority at The Times. As Ms. Abramson told me, it’s unrealistic to preserve an “immutable, permanent record of everything we have done.”
I can't speak for her true motives or intentions...but part of the problem here is technical ignorance. Her perception seems to have been that tracking changes would have been too much of a burden -- and she is justified in thinking this if she has never worked with something that uses a diff-like program...and in my experience, this includes most news people...in a newsroom, editing and re-editing stories and supplying reasons (i.e. as in a correction) is a very manual process. If a newsroom is using WordPress, sure, they know of the concept of article versions...but they don't really grok the ability for machines to create timestamped diffs.
However, if Ms. Abramson was familiar with the kind of system that Github (or Wikipedia, for that matter) uses, she would see that keeping an immutable, permanent record is actually quite easy for machines to do. So to create a inhouse change-log-per-article would be quite easy, technically, and with some interface work, it'd be a very welcome feature to more than a few readers, IMO.
edit: In terms of "most news people" not knowing diff...I'm basing this on anecdotal experience with non-programming investigative journalists who, among all journalists, have the most to benefit from understanding the implications of an efficient document diff, but I haven't worked with any who demonstrated that grasp. A common scenario is having two versions of some important document and wanting to see every change made. A diff would not be perfect, but it doesn't have to be, it just needs to triage possible changes to further investigate.