One can accept the core of Wittgenstein's perspective and the notion of language as usage without rejecting the notion that there may be a proper domain for metaphysics. No one is correct about everything, and people are often most blind to the consequences of their own favorite observations.
I would contend that within the theory of Wittgenstein itself there is a proper domain of metaphysical inquiry. The theory arranges language into a set of overlapping 'games', each of which is coped to best cover a given set of uses.
But any concept has its meta-version. If there is a domain of games, there is a domain of studying games themselves, which for Wittgenstein was the domain of philosophy. Philosophers of Science, for instance are interested in the 'game' of science, and what its proper rules are.
But there are questions about games themselves. First and foremost is the question of the boundaries between games, and how to negotiate them. When neurology and psychology both try to give interpretations of depression, there is a boundary dispute. Each considers the other's explanations true in their own domain, but somehow 'cheating'. If I drug you into being happy, have I really helped you improve your life, well, yes and no... If I talk you out of being depressed, but your physiology makes it likely you will fall back in within the next five years, have I really helped -- well yes and no... Yet, by and large, there is peace between these two cabals of cheaters, and each can incorporate the other's work despite its origin within a field with different standards.
So to the extent different sciences are sub-games, there are boundaries between sub-games, and ways of negotiating those boundaries that respect both of the adjacent disciplines. The same thing obviously happens between top-level games, like religion and science, or logic and physics. Decisions about how one is to resolve those are going to lie in another game entirely, a meta-game that clearly has rules of its own.
It seems to me, without forcing the perspective, that, other than the sheer volume of absolute nonsense that Wittgenstein was so appalled by, this is what metaphysicians have always done, and will need to continue doing, despite the dismissal of them.
I do not consider the author of the "Prolegomena to Any Future Metaphysics that ..." hostile to our continuing to develop metaphysics. So I feel your characterisation is an unfair oversimplification.
Kant rules out a certain kind of certainty entirely with the split between noumenon and phenomenon.
But he also clearly thinks he can properly talk about ideas like Categories and Duty, which underly certain phenomena and are to a very high degree absolutes. So we can get around Kant's problem with metaphysics the way he did, by properly understanding what he meant.