When I studied Hilbert space as applied to differential equations, I saw a perspective in which boundary conditions constrained the solutions in the same way as the differential equations themselves.
Physics has benefited in seeking simple differential equations—as simple as possible, but no simpler.
If this is applied to boundary conditions, such as initial conditions, then the uniformity argument for inflation is removed.
I have debugged numerical solutions for several sorts of partial differential equations.
I always start with uniform initial conditions where I know the answer and find the majority of my bugs there.
I am thus not surprised to hear that our universe seems to have started smoothly; I would have expected as much.
I recognize that this is a week form of argument, but it feels to me as strong as the arguments for inflation.