Former Teach For America high school computer science and math teacher here. (I also taught at a school funded by the Bill & Melinda Gates Foundation's High Tech High initiative noted in the summary.)
First, some positive comments. It's great to see studies like those mentioned in the Newsweek article attracting eyeballs in academia and the popular press. The conclusions may seem to border on the tautological for most of us (great teachers are great at teaching!), but such ideas are largely verboten in the public school system. If you haven't already RTFA, I'd suggest The Atlantic's treatment of the same material.
Anecdotally, I can fully corroborate Teach For America's data. Both in my school as well as those of my TFA colleagues, teachers that continually pushed themselves to excel and improve in their craft were able to consistently produce jaw-dropping results in their students' test scores. It really is amazing. As an example, I co-taught a summer school pre-calculus class with another TFAer in Watts a few years ago. We somehow managed to march through three years worth of material in those two months; our students went from being on average two grade levels behind to slightly above grade level. I attribute this success to Teach For America's philosophy of teacher excellence (which is similar to 'kaizen' in many regards).
The summary asks "What makes a good teacher?" This is the wrong question. There is no one thing that will make a teacher great (vibrant personality, deep subject knowledge, an M.S. Ed., etc.). Rather, it is an attitude that is willing to try anything (and, conversely, promptly reject the ineffective) to make students succeed. To use a math analogy, it is the second derivative that matters, not the current value or even the slope.
Disclaimer: this post does not necessarily reflect the views of my former employers.
Statistics are no substitute for judgement. -- Henry Clay