[bzbarsky]

This is only true for objects that have a spherically symmetrical mass distribution and for a point test object. That is, the field of a spherically symmetric object is the same as the field of a point mass of its mass located at its center.

As soon as your test object is non-point you get tidal forces due to the field typically being nonuniform. As soon as your object is not spherically symmetric you get a field that looks nothing like that of a point mass.

[starwed]

It's still the case that if you're sufficiently far away from a nonspherical object, it'll effectively act like a point source.

I'd guess this is what the OP was recalling.

[bzbarsky]

Well, sure. If you're far enough away it _looks_ like a point source!

That doesn't apply to being on the face of a cube, though; that's not nearly far enough away, obviously. :)