I'm currently using both Ultimate Terrain X USA (latest version) and FS Genesis 10m USA mesh V2. Even though both products use the most accurate data available, small roads, and especially streams, don't align with the mesh.For example, you might see a natural groove in the elevation mesh, but the stream is not following this channel as expected - instead it's slightly misaligned, or it criss-crosses the groove, constantly climbing and descending the stream banks etc. With the standard mesh, this problem isn't as apparent, but that might simply be because the 38m mesh isn't detailed enough to show small depressions and channels formed by streams.The same phenomenon can be seen with minor roads in mountainous areas sometimes. A road might cut into a cliff when it's obvious that it should follow the base of the cliff instead. This of course looks ridiculous.Will there ever be a solution to these problems?

I'm currently using both Ultimate Terrain X USA (latest version) and FS Genesis 10m USA mesh V2. Even though both products use the most accurate data available, small roads, and especially streams, don't align with the mesh.For example, you might see a natural groove in the elevation mesh, but the stream is not following this channel as expected - instead it's slightly misaligned, or it criss-crosses the groove, constantly climbing and descending the stream banks etc. With the standard mesh, this problem isn't as apparent, but that might simply be because the 38m mesh isn't detailed enough to show small depressions and channels formed by streams.The same phenomenon can be seen with minor roads in mountainous areas sometimes. A road might cut into a cliff when it's obvious that it should follow the base of the cliff instead. This of course looks ridiculous.Will there ever be a solution to these problems?

In my opinion the solution will occur when FS starts using photo based (such as google earth) scenery as a standard. Databases of points using limited generic textures trying to convey real scenery can only go so far-and it appears you are seeing these present limits in terms of rivers and roads. There also are many other present limits in the current method such as lack of subtle coloration of terrain, natural landmarks, and a almost limitless variation of terrain, soil type, coloration, and vegetation. To me we are facing the same transition we took 10 years ago when line based scenery was changed to the present database driven scenery(Pro Pilot was the first to do this). Imho we are now ready for the next step, as databased generic scenery with a limited scope can never match real scenery with all of its almost infinite variations.