> we also don't know if the space time on those scales is "smooth", if it isn't then light will not travel in a straight line which would account for the lensing effects.
Except space isn't perfectly smooth. It does curve, and that curvature is gravity. That's what general relativity is all about. If you want to postulate some way for the universe to curve absent observable mass, go ahead: you're doing nothing but restating the dark matter problem in a (minimally) different way.
> If you want to postulate some way for the universe to curve absent observable mass
The de Sitter vacuum is an exact solution to the Einstein Field Equations of General Relativity that does exactly that.
Our expanding universe will very likely closely approximate de Sitter space in the extreme far future.
There are many other vacuum solutions; it's normal in General Relativity to think of curvature as being a background laid down by hand (ideally approximating nature) and then perturbing that background with matter or probing it with test particles.
> Except space isn't perfectly smooth.
General Relativity is defined on a smooth (infinitely differentiable) manifold.
There can be arbitrary curvature in the manifold, however, and unfortunately with the Einstein Field Equations you can pretty easily produce manifolds with curvature singularities, and those are manifestly not smooth (and that is one driver for research aiming to extend General Relativity).
But to the first order, your experience of gravitation near the surface of the Earth emerges from the very weak manifold curvature the planet and its microscopic components and their motions source. The nearest strong curvature is a couple of kiloparsecs away.
Except space isn't perfectly smooth. It does curve, and that curvature is gravity. That's what general relativity is all about. If you want to postulate some way for the universe to curve absent observable mass, go ahead: you're doing nothing but restating the dark matter problem in a (minimally) different way.