DSM-5 also forgoes Asperger's and just includes it into autism now, which a _lot_ of people with either diagnosis tend to disagree with. There are many different ways autism and its variations present, and trying to "homogenize" them all into a single narrow label masks the complex nature of the condition (IMO).
There seems to be a ton of debate on lumping and splitting various disorders in the DSM. If I remember correctly academics like to split, practicing clinicians like to lump. There is so much overlap in many conditions that psychologists and psychiatrists that actually deal directly with patients question the need to split every small permutation of symptoms into a distinct disorder. I'm not saying the DSM-5 is absolutely free of issues as I'm sure DSM-6,7,8,20 will be an improvement.
That makes it a book in which some statements are social reflections rather than statements rooted in good science, which makes other such statements somewhat suspect.
It doesn't mean you should reject the whole thing out of hand but it might be a good idea to try to figure out how much of it is todays bias rather than good science in cases where that is warranted.
I'd contend that the most useful ground state for most of us, being people with no knowledge in the field whatsoever, is to assume that the DSM is broadly correct. From there you can, given some evidence, conclude that parts might not be.
I'm not sure, but I bet you're right. It is a medical reference that will change over time as better information becomes available. It's still the definitive manual for classifying mental and personality disorders.
That's political. Of course it's a disorder in the sense that it causes abnormal poor performance (reproduction is harder!). Peadophilia is also still called a disorder but only if it negatively affects the person's life. It hardly matters how we classify them. Everyone has some inferiority that they might want to improve, so we could all say we have disorders. Or we could all accept everyone as just-how-they-are.