Earlier quoted context omitted.
Without wishing to downplay the fact that this is a major issue, it’s hard to read scientific literature from a field you haven’t previously studied. Layman summaries are a great way to get a sense of what the paper is actually saying, though summaries made by random commenters are unreliable.
Many papers in general are just not well written, even if the subject is interesting. I wish it were seen as a social requirement that if you publish a paper, then you accompany it with a blog post—even if it's the only thing on your blog. With multiple authors, each author would write their own blog post. I'd rather read three different takes on the same subject, generally, than to slog through any given paper even…
While there are reasonably good reasons that most scientific articles are in PDF format (way too much work to change), I heartily agree with writing blog posts on articles. I started doing this: the effort (once you have somewhere to publish it) is negligible compared with the amount of effort of writing a paper, and it is also quite fun to be able to write more free-form text on the subject.