> [mammals] and chordata in general are highly diverse
Yes but we’re speaking of adaptability on the level of species. And the more specialised a species is to an environment, the higher its investment (in terms of some vague “evolvability resource”, which we might naïvely equate with number of mutations) in one particular strategy. This pays off as long as the ecological niche exists. Changes to that niche may require (relatively) fast adaptive changes, and these are in principle easier for generalists than for specialists. This isn’t a hard and fast rule, of course, but it seems to be a robust general principle. It’s therefore not correct to say that
> Specialization and [adaptability] are two sides of the same coin
Adaptability describes the ability of a species to adapt to new conditions; specialisation describes the state after adaptation has happened. Roughly, a highly adaptable species is a generalist whereas a highly specialised species is, well, a specialist: in this sense, the terms are (at least to some extent) opposites.