Not sure exactly what the other fellow is talking about, but science is definitely relevant here. If you read the recent decision in the Michigan Marriage Amendment case:
The basic question was: is there a rational basis to deny gay couples the same right straight couples have? One of the offered reasons for preventing gay marriage is that it was worse for children. That being a factual question, social science research was relevant. The short version: the judge, like other judges, found that the evidence offered favored the "gay people are just fine" side of things.
Indeed, what is better for children is a factual question, and the answer is not surprising at all. Of course gay parents don't matter, because parents don't matter. If you do literature search on the matter you will find long term effect of parenting to children outcome is (not statistically distinct from) zero when you substract genetic component.
http://www.freep.com/assets/freep/pdf/C4220110321.PDF
The basic question was: is there a rational basis to deny gay couples the same right straight couples have? One of the offered reasons for preventing gay marriage is that it was worse for children. That being a factual question, social science research was relevant. The short version: the judge, like other judges, found that the evidence offered favored the "gay people are just fine" side of things.