There's some different schools of thought on this.
A lot of content marketers and SEO experts recommend lots of articles targeting long tail keywords. However, the newer thought process is to write longer more in-depth articles and spend equal time promoting those articles.
As you can see from the article we published, we found longer articles lead to more shares. Brian Dean found that content length impacts Google rankings (http://backlinko.com/google-ranking-factors) and KissMetrics got a similar result as us in terms of shares being correlated with longer articles (https://blog.kissmetrics.com/share-on-social-media/).
The sweet spot for the length depends on sharing source and the domain you are writing about.
Thanks for your question.
More often than not, sharing long article is a form of self-praise (for me). I feel very pleased with myself sharing a piece from Aeon on Facebook, amidst cats and babies and other stuff: "Look at me, I'm deep and I'm making Facebook a smarter place".
So, it makes total sense that long read gets more shares per view, but it doesn't necessarily mean deeper "engagement".
So what you suggest could be true, the length of an article that does well on Facebook is shorter than LinkedIn and I'd argue that an article on Facebook is likely for a more casual reader than on LinkedIn.
However, even 2000 words is a fairly long article. They also found that articles under 500 words perform the worst.