This study is just way, way too small to have the conclusion it claims to have.
Whether left-leaning or right-leaning results turn up depends upon a wide variety of factors, not least of which is wording (left-leaning sites and right-leaning sites often use slightly different wording for the same issue). I don't see how it is even remotely possible for a group of four people to review enough searches to make up a representative sample, and four people is too small a number to provide a solid opinion.
Also, there's a distinct possibility that for some issues, left-leaning sites have talked about those issues far more frequently and in far more depth, while right-leaning sites will have discussed other, different issues.
Finally, I see no reason why Google should be held to any sort of standard of false balance: there are many issues for which the facts very solidly support the left, and it makes sense for Google's search results to reflect that bias. One of the clearest examples here would be global warming, but there are many others.