Also Unitys shadow mapping is by no means perfect and the Ogre implementation is also not that awsome if you ask me. It is just that shadow mapping for the sun is not easy to archieve at a high quality and good performence. I had something working quite nice already, which thanks to jcl now also works without an engine extension and with view frustum culling. I just got a bit away from working on it due to other projects. But I think that it just needs one problem to be fixed until it is well usable. Unhappily I already messed with that problem for quite some time now and still don´t really know how to solve it.

The great thing about gamestudio is that you have everything you need to realize any feature you want as easy as it could be. I am pretty sure, that even with plugins you are more restricted in unity. Unity is just much more high level than gamestudio.

I am btw just trying to show you that gamestudio isn´t bad at all. It doesn´t mean that I consider Gamestudio to be the only engine everyone should use. I myself have already worked with A5/A6/A7, Irrlicht, Horde3D, ShiVa, Löve and will soon get into Unity. I just think that every engine has strength and weakneses and that one can´t say that gamestudio is overpriced crap while unity is god.

If you look at unities "biggest update ever" features, you may notice that from the unity side most of them meant nearly no work. There is absolutely no problem to use that lightmapper with gamestudio, same for the culling, deferred lighting and g-buffer means basicly the same and just needs multiple render targets to be fast and breakpoints should be a standart. From a coding side, there is no problem with that stuff in any other engine. The difference is that unity can actually pay it. And from the view of someone who wants to create a great game that update of course IS big and I mean really big.

I think that gamestudio will survive on the beginner market for quite some more time, at least if they really work on the templates, what conitec seems to do currently.
Another possibility to survive is in my opinion the way the C4 engine goes, which convinces with hightech and probably a flexible programming interface.
I see the biggest problem for T3D, but maybe just because I somehow always kinda disliked Torque...