We're not talking about a fully-themed level. We're talking about using a terrain generator to get some hills, filling it with grass with a realistic grass rendering thingy (sorcery), and putting a couple of loops here and there to play with the physics a bit.
sorry, I was speaking of the game in general not terrain,
as when I saw things like adding little planet into the background and doing a sonic CD inspired environment, ect ect,
just looking at stuff like that, which got me a bit worried about the project's future of being organized.
I think my mentioned post about stuff referring to scales ect should be done first before starting any type of level building,
as once the scales are set than they will most likely cause the project to be stuck with them later down the road.
(You gata walk before you can run)
This being a graphical test, I think it would be waste of valuable time for the team to work on a level such as that to test what the engine can do.
Keep it simple, for a grass / terrain test you don't need loops, as its just a grass / terrain test, ( unless the loop had grass on it)
and for testing other shaders,,, they would most quickly be done on test object placed around a small room environment. (spheres)
after all the initial testing is done, than a level can be built quicker using all the assets you've created.
(or even have a separate team build (or model) a showcase level, while others concentrate on the graphical programing aspects)
Take it from me working on Sonic CD Remix and AXSX since 2006.
The more ambitious tests like this become, the more that they will take away from other important aspects of developing.
For example, Many tests that I've done became way to ambitious, taking away time for building real levels.
Edit: crap , the topic is build me a test level.. so that what I initially responded to.
Edit 2, adding a couple of loops here and there wont be enough to test the physics. ( especially or the 2.5D stuff)
thats why I mentioned making level assets for testing thinks like physics, and setting a scale in stone to follow the rest of the project)
Edit 3: Personally I don't think a terrain would be a good way to initially build or test the gameplay engine off of.
as its topography is very chaotic and not set up as very well controlled environment for testing and troubleshooting things like physics.