The tables I've seen all reference one or more studies, and while you may find that they aren't all identical, they're generally pretty close. You just have to read closely to make sure you're looking at the right thing -- some of them are designed as references when the gestational age is "known" while others may use an estimated gestational age, etc. However, I suggest that in many, if not most, instances the gestational age is an estimation at best. Do you feel you "know" the gestational age and you're trying to compare against "normal" weights, or are you trying to determine gestational age by looking at the weights/measurements? One would think it shouldn't matter, but there are slightly different tables specifically addressing those two different issues. Of course, as a fetus, there's also the issue of when fetal death occurred, and there's a decent series of articles addressing that...from 1992, but I don't know a more recent better series.
I'll have to look at some of my reference texts next time I'm at work. Personally I use the tables in the latest 2 volume Potter (and if I'm generating an autopsy report using that information, I typically reference the table &/or photocopy it including reference information and include it in the folder); that doesn't mean there aren't others out there, but it's broadly available, broadly accepted, referenced, and thus defensible.
(I'll take this moment to plug Zotero as I went to it a couple of times while writing this, and have enjoyed it for several years now. Zotero is a free Firefox plugin which serves as a bibliographic/reference management database, including full-text attachments, and can be used with Word or OpenOffice when preparing manuscripts with references for publication. A brief example of use is when I'm using PubMed and find a useful article, a single click imports the info + abstract to Zotero, and another couple of clicks adds the PDF if available and opened with Firefox. Compare to EndNote and the like.)
