1) In fact, you cannot ādamageā your system more from running the build program than, for example, from C. On a modern OS, the worst thing you usually end up with is the failure of your process. However, it may be a little easier to compromise your process from assembler code. If you want to get something useful from the emulator, check out Bochs or QEMU - instead of just providing a sandbox, they have useful debugging tools. You will be fine without vm, though by writing + test code on your regular OS with your regular debugger.
2) Actually not. You can do a transfer between the OS (but with the same CPU) as you do for C, but building cross-cpu just doesn't make sense (if you want to do this, you can use C or .NET or Java byte- the code).
3) The theme of many fire. Depends on how you look at it ... Assembler is just a compiler that deals with assembly language. There are many tastes of the assembly, some above others; most of those available for x86 support macros. HLA adds a lot of macro support and syntax that many don't like, plus a large standard library. Not my cup of tea (I would prefer to use the "real" HLL), but in my book it is still assembly language.
4) http://www.asmcommunity.net/board/ , http://board.flatassembler.net :) - stay away from IRC channels. Also, stay away from usenet alt.lang.assembly, it was almost destroyed by several middle persons.
As for a few reasons to study the assembly:
6.If I learn assembly, getting to know HLL will be easier.
Not necessarily, and it can be even harder if you "do too much" before breaking into the HLL. I find that some knowledge of machine architecture can lead to writing better software, but only if you learn to constantly not dwell on tiny irrelevant details.
7. You have something to offer when you apply for a job that others will not have.
It will not necessarily help you, since the assembly is absolutely irrelevant in most positions. Damn, you risk the prejudice, "oh, he will never achieve something lost in worthless microoptimization."