Unfortunately this is what is possible to render with only celestial coordinates as input data. I'm looking into getting redshift at data right now, so that I could have the camera at the center of the earth and looking out from earth at different galaxies with different distances and brightness.
Given that oftentimes you get a range of estimated distances (in megaparsecs, MPc), maybe you can show the galaxy as a line instead of a point. The innermost end of the line would be the closest estimated distance. The furthest out end of the line would be the furthest estimated distance. You can put the median estimated distance as a dot on the line. Either that or you'd have to use the median estimated distance for a singular point.
Andromeda we have pretty-well established at 0.78 MPc, and M32 at 0.77 MPc.
NGC 7768 at 120 MPc.
Whereas the largest galaxy we currently know of, ESO 383-G 076, is at 200.59 ± 14.12 MPc. That gives a variability in distance far larger than the distance to most of our nearby closest galaxies!
Regarding point 3: You are correct! the galaxies are in fact clustered together in space. These groups are called gravitationally bound clusters.
Best regards,
Victor