In Newtonian mechanics, the gravitational potential can be anything you want it to be, since it's arbitrary up to an additive constant. It isn't a physical substance that exists in the space between two galaxies. If you want to estimate it relative to its value within a galaxy, just plug into the usual equation, approximating the galaxy as a sphere.
In general relativity, there is not any way of defining a gravitational potential that applies to cosmological spacetimes. That means that if you try to generalize this to discuss things happening on cosmological scales, it won't work. The Newtonian approximation only works for your original question because you were describing something within a cluster of galaxies, which is much smaller than the Hubble scale.