These are good steps; obviously more needs to be done, but there have been some quantifiable improvements, both in the metrics cited here and in my OS observations blog.twitter.com/official/en_us…
However, in my observations, these measures have been hitting a lot of low-hanging fruit in terms of obvious commercial spam and bots. The more insidious influence activity has adapted, becoming more sophisticated ahead of these measures.
I'll have something out probably next month that looks at some of the latest technical tells. These aren't rocket science, but they do point to a next-generation approach to automated tweeting.
