Considering hyperlexia as a separate disorder isn’t a problem, though.
Alexithymia also used to be thought of as a part of autism, but now it’s recognised as highly co-morbid, i.e. it’s often seen alongside autism, but not exclusively. It’s possible to be alexithymic but not autistic and autistic but not alexithymic, in much the same way that Brooke described hyperlexia.
Separating both alexithymia and hyperlexia from autism but noting that they are highly co-morbid with it, in the DSM, is good for all affected neurodiverse individuals (and their parents), from the point of view of getting the right kind of support.
I expect that autism will eventually be seen as a highly correlated cluster of traits that all arise via similar causal mechanisms, such as heightened testosterone levels in the womb at certain times during gestation interacting with genetic predispositions. The point being, the features in the cluster are highly correlated and suggests a cascade of some kind, but not every child is impacted by the whole cascade for a number of reasons (genetic predisposition potentially being foremost).