Well that all escalated quickly after my post
Maybe if I post again it'll all stop.
So yeah on grass, I've gone down this pontification loop before and really there is no easy way to get the result you want unless you're willing to get in there and really think about grass like you'd paint grass in a nice landscape.
Is grass all the same length, type, straightness, colour etc? Nope.
Long grass bends over more, short is straight up but almost invisible.
Time of year will dictate the look too. Seed heads, no seed heads. Colouring. New shoots. That sub-concious look of 'that's not right' because you used autumn look grass with spring time tree look etc.
Eventually I just created a 2048x2048 texture for grass (could probably strip out the 2048 and run from the 1024 mip downwards with a 1024px texture)
It then had a whole atlas of long/short grasses, shrubby bits, couch grasses etc, and these all mapped to scaled planes for distribution, so they all looked about right scale wise and pixel density sizes.
Then I'd just paint them in with scatter tools. Time consuming yes, but also nice because you'd naturally paint more density near track edges, or slow areas, and less density back off towards the barriers or at faster areas.
I'm pretty sure back when I was discussing this with (Nighteye, iirc, back on the AC forums when you could), I'd said that I'd given up on scatters and just did it all manually because it just made the best result vs many iterations of scattering trying to get 'a look' that you'd never be 100% happy with.
BUT, all this is kinda pointless any way because AC's grass method is crap.
I think Dirt was it, or some Codemasters game, had textures for the track (full size over entire course), that'd procedurally place appropriate grass, and had a lower GPU and RAM load than millions of bloody 2d planes everywhere!
It wouldn't be so bad if there was a UV coord jiggle so the grass could wave, and it'd give tracks much more life... but no, even though Free Racer had that back in about 2006, and it looked pretty damn nice with good settings, KS decided to just use the method from 2004 for grass.
Personally I've done about 20 hobby tracks and finished maybe two or three of them. I've done about 20 pro tracks and done them all in quick time and they look fine with happy clients.
Pontificating and learning and trying out ideas is cool, but sometimes just moving on and getting the whole track kinda right, and then doing multiple passes afterwards to refine it is ok too.
I think my original Top Gear test track was re-spun loads of times in conversions etc. It was terrible really. It took like 5 days to make and involved pontification too haha. Maybe 3 days of actual work tops.
But it got out there and worked well enough and was enjoyed. It also gave a great perspective on what needed more work and what could be left alone.
All the best
Dave