Ripley, you really have no understanding of how things are rendered in a 3D environment.
Just real quick. There are images in this thread which could be rendered using two triangles and one bitmap and a handful (< 20) of point draws. That is how much work it is to display space. A typical frame of any ground based game is usually around 75,000 (or more) triangles, and many bitmaps (bump, normal, specular...).
High detailed terrain will always trump space based games in the effort it takes to draw them.
You're making the assumption that the planets etc. in the images are just flat 2D backgrounds which they partly probably are (level of detail adjust) but the planets and stars in the jump system are highly unlikely to be just 2D objects - not to mention other players and docking stations.
That was not the point however, the point was that game engine makers seem to create techniques like field of view blocking like in the image. I know for a fact that the more developed game engines use level of detail maps that reduce detail according to view distance (for the reasons mentioned before) and also draw only the visible area of the screen. This means that graphics wise it makes no difference what happens all around you, you can only see the fairly narrow strip that is your field of view.
That, then, means that map size plays no role whatsoever in obtainable graphics quality as far as GPU rendering power requirements go. Your system is either able to render that small strip to the longest set view distance (which is not the whole map border to border unless the map is very small) or not. Makes no difference if there are a gazillion objects that are outside your view distance or field of view. Well, unless you have an engine that renders invisible objects and bogs down your system due to not being optimized.
In AH that would basically mean that when you sit in the cockpit, nothing that is behind your cockpit (dash, bars, wings, clouds etc) at any given second, does not get rendered or even handled in the GPU untill the very second it crosses your field of view and needs to be displayed. Just like that field of view image, all the 'shadow' areas of the image do not even exist as far as GPU sees it.
I'm basing all this to the stuff I've read of the technical documentation of various game engines. Of course I might be totally wrong - this is the impression I've got from studying game engines so far. Maybe there's some specific reason why invisible objects should need to be rendered outside of what you can see.
I'm not taking into account the logistics of tracking hundreds of players and objects which is obviously huge, I'm only talking about the rendering part.