Questions about torque

Hola everybody! i'm new to the garage games scene and the first thing id like to do is thank the founders for creating and offering such a cool engine at a reasonable price. its also cool to see a community of very nice and helpful members providing valuable resources and valuable feedback. i love this place so get used to this mug! ive never made any games and have very little experience in graphics programming but I'm learning more and more every day and I have some great ideas for games and whatnot. ive already decided to use torque for all of my games. although I havent officially purchased the engine yet, i plan to and I've been studying my copy of 3d programming AIO closely in preparation for the big event.

anyway, onto the questions:

first off, are there big differences between the book version and the regular version of the engine? i understand that the book version is a compiled version of the engine, but i have no idea what the engine is like in its pre-compiled state. my main concern is having wasted alot of time studying functions and script layouts which are particular to the book version. i guess what im asking is: is the common code base and callback functions the same in the book version of the engine and the straight out the box compiled version of the engine? is there alot of legwork involved in getting the engine ready for its first compilation or is it ready upon arrival?

second, is the engine highly customizable? the reason i ask is because i used another engine earlier called panard vision. its a great engine which came with source code, the only problem was that the real meat of the engine was hidden from the user in a dll. would i be able to configure the engine at the polygon or rendering level? say i wanted to tweak the graphics pipeline. is the engine code configurable at this level?

third, what does torque use for hidden surface removal? i dont mean dts or dif objects, im talking at the world level. I know the terrain engine uses geometric LOD which is a very important, but does the terrain engine feature occlusion culling of hidden surfaces, poly sorting, zbuffer, pvs or what? does the algorithm work on the heightmap? im sure I could figure all these things out if I just actually took a look at the damn thing but I have been wondering about this for some time now and whether I buy the engine may depend on the answer to this question.

oh and one more thing, is the code source commented? i know the most important thing is that it works but its much easier to understand and modify code thats documented to you can know whats going on. and does it come with comprehensive documentation? I dont just mean enough to get me started, does it come with a complete, organized function and global reference?

thanks for whoever actually took the time to read my post. any information one of you guys can give me to help out or point me to where i can get this information would be greatly appreciated. and yes I always type with no capitals so get used to it!