I went through the top results of that search, and most of them make no claims at all about the performance differences of diverse vs non-diverse groups (since they instead adress some quite different questions), and those in the top results of this search who do look into performance comparisons do not find such an advantage, they find that there is no significant difference. Furthermore, a large part of the studies are focused on artificial short-term cooperations e.g. study groups or team projects at a university course (which actually eliminates a key part of diversity by ensuring that all the team members are of a similar age group and of similar specialization, they essentially only measure the effect of ethnic variation) which aren't really representative of what the success criteria are in a business environment.
Perhaps there are studies that would justify a claim that "diverse teams are more successful", however, this particular search does not immediately lead to them and should not be used as evidence of some scientific consensus on the matter - quite the contrary.
I think it's likely that certain types of diversity produce better outcomes. My problem with the research is publication bias, publication bias where finding the opposite will cost you a career, most of the experiments being contrived, social science being a joke field, and the sort of people who are attracted to social science.
Perhaps there are studies that would justify a claim that "diverse teams are more successful", however, this particular search does not immediately lead to them and should not be used as evidence of some scientific consensus on the matter - quite the contrary.