What could effectively solve this is if HN attached a short abstract to each article, written and updated by user submission. (The initial version could/should be provided by the submitter.) Maybe minimal GitHub-like version control with pull requests accepted by vote. This could be useful not just to save time, but also to keep article quality higher, if submitters are required to justify the intellectual value of the submission. The abstract would be more valuable than normal comments, so accepted modifications could get extra karma.
This would have similarities to summaries on slashdot (which are written by moderators and so not very scalable) and the "submission statement" required in certain subreddits.
https://www.reddit.com/r/geopolitics/comments/3yh1ll/introdu...
https://www.reddit.com/r/TrueReddit/comments/1p4gqg/new_poli...
https://en.wikipedia.org/wiki/Inverted_pyramid
It's really about news vs. human interest pieces, and many outlets (like the NY Times) feature both. Seems possible to me that American outlets produce more of these, but it'd be hard to disentangle this from the general dominance of American outlets on the English-speaking internet, such that non-American pieces are more likely to be directly local and, so, news rather than human interest.
"People will read 3 words to decide if the topic is interesting. If they think it's worth their time, you have thirty words to hook them. If they take the hook, you can probably get them to read 300 words unless they have a focused interest in the topic."
It is, unfortunately, tied to much evil online (the quick-list ad blog ecosystem is basically application of 3-30-300 writ large), but useful for creating general-consumption articles.
Long-form science writing isn't inferior to news briefs or press release-style articles you're looking for; it can be a very effective method for engaging a reader and introducing them to a topic which they don't have detailed knowledge on (or which they just want to read about).