To summarize, in my opinion, the gist of those articles:
1. Consider your niche
For example, when I'm working on a programming project, and I'm googling an error code, or question like ' how to do [task] in [language] ', I almost always sort the results by most-recent, as language updates often render older examples as irrelevant or at the least, problematic. If your niche were something where historical accuracy were relevant, then perhaps an older date would actually benefit you!
2. People like New Content
I think the shoutmelound.com link and moz.com links pretty well illustrate that just showing dates doesn't help you (it actually seemed to
drastically hurt the shoutmeloud.com site) but showing updated dates with a closer to current date seemed to help moz.com in their test. My conclusion from that is that people view newer content as being more valuable. If you show them two articles, published on the same day, one year ago, and don't tell them the date either one was published; they'll likely read each in a more objective manner. If, however, one article is updated to display a date as being more recent, it would seem more people will choose that article. I don't think this is an odd phenomenon, and simply reflective that newer information is more appealing.
3. Update Your Old Content
Tweak it, fix links, make the pubDate the current date. I'd like to see some data (I will be doing my own tests too) regarding the effectiveness of the level of update. For instance, if I built a script that simply updated the published date of all my content once a month, would that give me a bump? If it didn't effect my SERP rank, would it still benefit my CTR in the SERP simply because my published date was newest? Obviously, it'd be best to take the advice in the ahrefs article in adding new content, updating outdated info, fixing links, and building new links, but I'm sure there's a nice pocket between too-involved and effortless that produces a nice ROI for time vs. benefit.