These kind of studies tend to be riddled with methological errors. For instance, one of them found (as usual) that autumn-borns had lower grades than spring-borns. However, once adjusting for the parent's education level, the results were opposite.
If "doing the best we can" is too low of a standard for research, then it's all garbage. Yes, there are many problems with some of the research out there, but you deal with it, not ignore it all. The purpose is to reduce uncertainty, not "get the right answer".
Edit: Not to mention that if you want to publish "tradition", don't call it "science".