I suspect that their new formula for matches which they say includes "well researched trees" may include adding people to circles who match on the DNA but do not have their ancestor documented as far back as those "well researched trees." This may be a good thing IF the DNA of everyone in the circle is the same place on the same segment AND the trees they use are indeed well researched. So, yes, maybe helpful, but we STILL NEED a chromosome browser.
I now have some people in my Haigood circle who do not have a hint leaf, but do have Haigoods in their tree. I have traced their ancestor and can connect them to my family per other ancestry trees, they just do not have the information yet. I have gotten several of my Scotts and Mauldins to add the ancestors they are missing and have informed others but either they will not add the information or they do not put enough detail for Ancestry to make the match. I have documentation that they do belong to my family.
This could also be a BAD thing. There is a very common error made in my Mauldin line to connect a man named Henry Laban Mauldin as an ancestor. He has about 20 children connected to him in over a thousand trees, BUT he was not the father of all these people. This is proven by DNA. In fact he appears to be a fictional person concocted from combining records for a Henry Mauldin and a Laban Mauldin and most of the children attached to him do not belong to Henry or to Laban. So these trees with incorrect information may be coming up as matches or in circles when they have totally incorrect information. The DNA matches would appear to validate this incorrect ancestor.