This was a study of some 20,000 people with T1 who had registered with the T1DExchange. This was the first study result I have seen from this effort. It is interesting to note that the authors report an association with high statistical confidence between frequent testing and better control (HbA1c), but in the abstract they don't report how much better control. This usually indicates it was not a particularly striking difference.
The study also reported a strong association between frequent testing and non-Hispanic white race, insurance coverage, higher household income, and use of an insulin pump. Thankfully they didn't pull the typical bonehead move of concluding that frequent testing can make you non-Hispanic, provide you with insurance coverage and a high paying job as well as making an insulin pump appear out of nowhere and attach to your body.