I would like to simulate an non-neutral system of an infinite dilution electrolyte solution - 1 Na+ ion in ~500 water. Are there certain parameters and settings I need to use in the input file to account for the non-neutrality? Do I need to fix certain kspace settings?

I vaguely got the impression that you can use MSM for non-neutral
systems, but I'm too lazy to read the Hardy paper.
Hope these links help in some way.

I would like to simulate an non-neutral system of an infinite dilution
electrolyte solution - 1 Na+ ion in ~500 water. Are there certain
parameters and settings I need to use in the input file to account for the
non-neutrality? Do I need to fix certain kspace settings?

no. the non-neutral nature of the system will cause the kspace sum to
converge, but that convergent term will be ignored, resulting in a
behavior, as if the system was neutralized by a counter charge spread
evenly arcross the surrounding environment.

for a system like yours, this should result in an acceptable small error.
you will still have to deal with finite size effects (your system is not
really infinitely diluted), but also a similar size and neutral bulk water
system will have finite size effects. as an alternative, you could consider
MSM (as already mentioned, but that is difficult to converge to accurate
forces) or one of the approximate short-range only algorithms, like DSF or
Wolf. as you may see, nothing is a perfect choice. it is advisable to
carefully study the published literature on such systems (i know there are
publications on similar systems going back over 20 years). i am certain,
there will be comments on accuracy of the different approaches.

of course, you are also encouraged to make your own tests. 500 water
molecules is not very large and it is usually easy to get access to CPU
time, so you could try running larger systems and compare results and see,
which method gives sufficiently converged and accurate results.

sorry, correction. of course, i am talking about divergence not convergence. ewald summation will simulate and infinitely large crystal and an infinitely crystal with non-neutral unit cells will have an infinite energy.