As for development platforms, I'd like to see it made in the torque game builder myself, as you said it would probably be easier for people to jump from GM or MMF.
This right here is my primary reason for wanting to use it. As well, it provides support for Mac and Linux platforms, so our non-Windows developers (though they seem to be few and far between, here) would not be left in the cold.
Otherwise, I think C++ and SDL, they would be a little harder to grasp but they're a lot more flexible.
The issue that we run into there, however, is that the tutorial would start mostly on how to design a game engine from scratch using those, vice how to design a zelda game, though that would be alleviated by me developing a custom engine around SDL.
I reckon torque. Mind you, you'd have to get these tutorials finished in 30 days, or people will have to "obtain" a full copy of it 
.
That was the biggest issue that I ran into when considering the use of Torque as a development platform.
Also, I think that going for the latter option of "how to design a zelda like adventure game" is the better way to go, because people can branch out later.
My thought, exactly.
C# simply because its simple and easy for people new people too pickup. It also has more future opportunities than say Torque.
While I know that to be true, and I love C# as a language, there are some things that need to be considered here. Firstly, not everybody on ZFGC is ready to jump into a full development language. Beside that, the past two tutorial runs have been using C# in one form or another, and have not been very active, probably for that reason.
Second, I would venture to say that we have more people here who wish to become designers and artists, vice programmers. Using a proven technology such as Torque would actually provide them with
more in terms of getting a job as a designer than they would receive if they had stuck with C#, in which they would be limited by how well they can learn and adapt their ideas and desires into a programming language. Logical design, such as setting the flow of a game engine for optimal performance while achieving the desired effect, is very different from, for example, visual design.
Custom Lib because you can dictate exactly how much depth you want to go into with the tutorials (for example you could wrap all rasterization into the lib, all pak files, etc, etc, and only concentrate on higher level game design specific aspects). You could also leave the library open source so those who wish to progress further can then study how it works.
See above.
As for using other libraries; you then have to explain to everyone the processes of linking them all together (and fixing all the errors thats going to undoubtably cause u_u), you also have to explain how each one works, and teach them the syntax/naming-convention each one uses, not to mention there are few that work perfectly with C#, most are in C++ or similar.
It would not be very different from explaining how to link your own library into the project. Any library that I use for this situation will be similar in syntax and in how it is used, as with the various SDL libraries.
