The simple answer is that while Diamond's conclusions may be flawed, he's not necessarily wrong about everything. So it's fine to cite a chapter of his work, when you reckon he's done the work properly in it, and it's supported by other sources.

Given that Diamond is, in that chapter, essentially summarising work in climatology and comparative botany, it's simply a handy place to point to, where the other option is a lot more legwork to bring together much the same research from other sources.

More broadly, it's useful to read the things that are deemed wrong as well as the things that are deemed correct, and make up your own mind on the matter. There are always scholars whose work is disputed for what seems like solid reasons at the time of publication, but who are later found to be more useful than their contemporaries, and vice versa. And reading things that are not in the mainstream of academic thought can bring up concepts, models and motivations that you wouldn't otherwise come across.