The only change made to the "cedar" was the density. The difference between that tree and the Lebanon cedar in terms of density is well within the natural variation, so if you're going that route, the only thing wrong is the comment, which has no in game effect at all. That is, of course, my fault, but I don't think it a major problem. That thread has been posted for some months now inviting comment and further research, after all, with that only being noticed last night. While you are correct that we had to take averages for many things, this is a limitation of DF. We simply cannot assign each stone some random density within the range at present. And even if Toady did that, all it would really do is to make DF more memory-hungry, because every one of the stones in your fortress would need another integer attached to it. That would be undesirable.
I'm certainly in favor of doing more science on the subject, but given that density is important for hauling speed and fall damage, both of which are important in the new release, I think that we need more science, not less. You say that you don't see the point of changing to arbitrary values, but the materials already had arbitrary default values. In the case of wood and many types of stone, these were based on the rock and wood templates, which were utterly out of line with the actual materials.
Put another way, the choice was between somewhat-realistic arbitrary values and completely unrealistic ones. Insofar as you can help make more realistic values, please do! I spent several months researching, but I'm no expert on trees. And trees probably do deserve more love in the raws. If you want to help, let's work on a list of representative world trees that should be in game, based on real species. Then we can split up things like oak into their subtypes, research proper growth locations, densities and colors, and mod this information in a way that allows Toady to incorporate it.