Please enhance long description

[sten0]

Continuing from #65

1. How does wgrep compare to the built-in replace-regexp and query-replace? eg: What is a 'grep buffer"?
2. Why would users want to use wgrep over the above methods?
3. Why would users want to use wgrep over helm-ag?

From what I can tell it seems to be like replace-regexp, except that it operates on all open buffers in an Emacs instance. Also (again, only from reading the description), it sounds like the "grep buffer" can be generated using a variety of backends. So I'm guessing is something that a) generates a pattern matching space using one of the backends b) that this then becomes an editable buffer c) finally there's an output stage that applies all changes to patterns in the "grep buffer" to all active buffers.

My feeling is that wgrep enhances efficiency in an interactive way :-) Please let me know if the above short paragraph is accurate so I can get started on a PR, unless you'd rather handle it yourself.

Thanks, Nicholas