Optimization is talked a lot about here, but what about using memory more aggressively for those of us that have large amounts of memory available (32 GB, 64GB, or more) ? DDR4 has made it very affordable to have such capacities ( I have 4x16GB) on recently built computers. Sins of a Solar Empire is a game that used an interesting technique. It loaded all of its assets into memory once at the start and had near instant loading screens as a result. Something similar might work well for TSW, Train Simulator of old certainly had a very noticeable hiccup on loading particularly detailed cells.
Unfortunately this would most likely increase the minimum specifications and narrow the number of people who could run the game and break it for a lot of existing players.
Hence "for those of us that have large amounts of memory available". Pre-cache resources if more memory is available, do not if there isn't any. Though I'm not familiar enough with UE4 to judge if resource handling system can be tinkered with to allow such behavior.
If you use the view distance increase hack, that is a way to massively increase quality of the scene simply by taking advantage of more RAM. DTG could go out of their way and make it so that us with large amounts of RAM could increase view distance longer than the currently allowed values and display some form of 'predicted RAM usage' on the draw distance slider. When something it an option, it doesn't matter what computer anyone has! But for now, if you want to take advantage of RAM, use the r.viewdistancescale tweak
That tweak doesn't really address MikeForce's point. It doesn't affect pre-caching resources from disk into spare RAM to minimize loading times and little hiccups upon loading new area. r.ViewDistanceScale affects distance at which already loaded and displayed objects are switching to lower LOD (LOD0 has the highest quality in UE4) or begin to display at all. Yes, higher parameter values utilize RAM a tiny bit more, but putting much more strain on CPU, GPU and VRAM. So this is a solution for different problem.