The idea that the lowest performing five or ten percent of teachers need to be fired on an ongoing basis strikes me as wrongheaded but at least has enough ruthless ambition to seem sort of plausible.
The idea (mostly from Gates minions) that the reason firing teachers might work soon (but not previously) is because we've never really known who the good teachers are or what they do is infuriating but at least has a certain logical consistency.
But the argument as Hanushek makes it today that a one-time purge of easily identified bad teachers would close the achievement gap between the US and higher performing countries like Canada and Finland is just laughable. If that were true there would not only be a lot of hard quantitative data already to back up the assumption, e.g., those reconstitutions in Chicago would have mostly worked, but there would be tons of anecdotal evidence to that effect. You simply don't hear about failing schools that have been turned around simply by getting rid of a handful of bad apples. Nobody even claims that with a straight face.
Also, I don't understand why this would let us pass Finland. Why would our current practices plus a purge be better than Finland. For that matter, why wouldn't we expect them to then adopt our method and increase their own scores in response?