just my personal thoughts inspired by this discussion about the 3dgs editors:

for me MED was a good start as a simple 3d modeller, but later found disturbing it because of its missing (nearly basic) functionalities. but after MED it is easy to get familiar with for example Milkshape, Fragmotion, Sculptris, Hexagon, which are cheap or free. moreover, some other engines do not have own modelling tools at all... as an engineer who worked for years with autocad and other really different cad/cam systems, 3dsmax is the more attracting one, but too complex and expensive for a hobby user (I used and liked gmax too for modding), I tried out Maya but found not for my taste, and Blender seems to be made on another planet (but would be really worthy to learn to use). for the first sight WED seemed to be a good and simple tool, easy to make simple levels from blocks, but after a short time I totally eliminated it from my workflow (because I need outer scenes with terrain, vegetation and building models). Terrain creation is also possible by some free/cheap tools (Terragen, L3DT, PnpTerrainCreator, Earthsculptor), the offered features are really basic in MED and WED. GED always crashed, so I gave up to use it at all.

so if I would own 3dgs, I would keep MED as is, with minor updates for catching beginners to offer a complete package (because advanced users can use a lot of other tools due to their taste), and would concentrate on WED to make it more up-to-date and cool looking, and a more integrated tool, suitable both for beginners and professionals - which is the hardest task. I like Lite-C, I would keep it too. but unfortunately the development speed is really not comparable to other engines at the moment...