No, similarity is GOOD when comparing policies. If I'm comparing how effective certain policies are in New Hampshire and Vermont, which have similar histories, similar cultures, similar demographics, and similar government systems, the comparison is a lot more valid because there is a lot less that needs to be controlled for. If I was to compare, say, certain policies of the United States versus France, the comparison would be much less valid since the two have different histories, different cultures, different demographics, different forms of government, and are located in different parts of the world. That's not to say that such comparisons would be totally invalid, but they would be much less valid than comparisons between more similar entities.
It is because the states are similar to each other that makes the comparisons valid. If Oregon and Washington implement radically different policies toward, say, crime, then we can compare the crime rates in the two states in a few years and see which policy was more effective.