'Either there is something wrong with the world, or there is something wrong with I.' — It is tempting to reduce this to, 'Either metaphysics is the knowledge of an ultimate reality, or metaphysics is a dialectic of illusion.' Either way, you have something called 'metaphysics'.
However, that's just equivocation. Either way, 'metaphysics is not about nothing' (see the article by Adebayo Ogungbure in Philosophy Pathways Issue 158). But we didn't just want metaphysics to be 'about something'. We wanted it to be about something ultimate.
'Well, that's too bad. This was about the search for truth. Whatever you find, wherever your investigations take you, that's what you have to believe.'
Of course. I accept that I have to go where the arguments take me. But why do I have to accept the alternative: 'either knowledge of ultimate reality, or a dialectic of illusion'? Kant didn't. In the Critique of Pure Reason he offers his own 'dialectic of illusion' in the 'Transcendental Dialectic'. But he also presents arguments for a metaphysical science of synthetic a priori propositions, transcendentally deduced from the conditions for the possibility of experience. (Then he hedges his bets and claims that there exists an ultimate reality of noumena or 'things in themselves', apart from the world of our possible experience, but we can have no knowledge of it.)
In other words, I don't have to accept the alternative, 'either a dialectic of illusion or a theory of ultimate reality', because there's an overlooked third possibility: a 'metaphysical science', a non-empirical theory about the world we know which is necessarily true, informative and substantial. Forget the stuff about noumena. But if Kant already did that (250 years ago!) what is there to add? Or did he do it? Was he successful?