The web has changed a whole lot. The original research link here is actually about two separate studies.  First, two professors looked at the most emailed articles on The New York Times site and found the most shared tend to be long form.  Second, SEOmoz, which I don’t know how they fair in your neck of the woods, but they have a credible reputation here, conducted a similar study of blog posts — and found the same correlation. Conclusive?  Of course not.  Compelling?  Yes, Sir. I’d never recommend someone set out to write 1,200 word posts when three paragraphs will do the job.  The length of copy is a lousy objective, but for a good writer with prose that flows — have at it. 

I think the things we agree on are:1) Long posts can work, as long as they meet other quality criteria and don’t waffle2) Short posts can also work, though there is a lot of superficial stuff out there.Personally I prefer to write more in-depth posts for the reasons we’ve covered.I think the only point of genuine disagreement here concerns the usefulness of the original research, but this may be one of those ‘intelligent minds may differ’ situations.

In either case, you are certainly welcome to your view and are entitled to write short posts. 

“Correlation isn’t causation (sorry, the phrase is cliché for a reason), and it’s possible that there’s something else at work here. Perhaps the journalists tend to write longer pieces when they’re writing on hot topics, for example.”

So talking about length in itself probably isn’t helpful.  Can you share the research backing up the shift away from Nielsen’s research?   Books come out of user-testing, such as ‘Letting Go of the Words’ by Jinny Reddish suggest that long-winded content is still a turn-off.