I use articles like the one Jes Scott posted as launching points for research. I expect both math and science news articles to be very general, as I do for Wikipedia, as I do for my nephew’s class materials.
If you take away the verbage, what you’re left with is links for further reading, links with authors.
From the referenced papers, I look to the author’s other work to get a sense of their specialities and I’ll read a few more approachable papers.
Once I have a sense of their specialities, only then do i go back and look at the complicated thing for now I have a footing.
In this case, it’s extremely easy. Only one man, Jacob Lurie.
That one man has research papers, many of which are here.
Only once I’ve done all of that can I poo poo research or praise it. I follow this process with any subject matter and it works for me. Stopping at an article and going no further makes no sense to me unless I’m going to critique the state of that particular publication or decry the failure of modern generalized math/science/etc reporting, which has sociological value of course.
Oh, I do love a good Press Release. Always “so excited to have you on our team”. It’s upbeat and positive.
Ah – what’s nice is that Jacob Lurie hangs around mathoverflow and answers questions about his work whenever it comes up – and given the apparently density of his work – it comes up a lot.