If by material, you mean less than d=0.3, then you appear to be correct.
It's gonna be really difficult to estimate reliable effects from a meta-analysis of such disparate studies, which is normally the case in psychology.
Anyway, it would make sense that cognitive interventions are more effective, as this would suggest that a lot of implicit biases represent pre-conscious reactions (which is interesting in and of itself).
Don't worry, if you read even the just article he linked, you'll realize that the quote he pulled is out of a list of multiple reasons their metastudy did not come up with conclusive results. There's a list of possibilities, and there not being a correlation is a possibility, but due to the nature of metastudies, they are not willing to make that statement.
It's gonna be really difficult to estimate reliable effects from a meta-analysis of such disparate studies, which is normally the case in psychology.
Anyway, it would make sense that cognitive interventions are more effective, as this would suggest that a lot of implicit biases represent pre-conscious reactions (which is interesting in and of itself).