Markdown and Restructured Text are pretty nice. One Markdown annoyance is having to use two spaces at the end of a line to indicate a line break. In some text editors, you can't see the spaces at the end of the line so when you visually scan your Markdown you don't see the line breaks.
One of the nice things about Markdown and reStructuredText is that there is actually a substantial overlap in their markup formatting - it is not too difficult to write documents in the common subset, e.g. this brief guide that I wrote up documenting the common subset: http://gist.github.com/1855764 (you can see the same file interpreted as Markdown and reStructuredText there).
My biggest complaint with the various lightweight/plaintext markup formats is the sheer variety and incompatibility of their markup syntax. Pandoc goes a long way to being able to convert the more popular ones easily, but it would have been nice if there had not been so much gratuitous re-invention of the wheel. It's kind of sad that the MultiMarkdown table format is incompatible with the tables in reStructuredText.
There are "extra" versions of Markdown that can be configured to respect literal line breaks. But if you rely on that, your text would not be compatible with other Markdown parsers.
Markdown's default behavior of disregarding line breaks has two benefits: (1) you can keep your lines short without breaking paragraphs apart, and (2) putting each sentence (or long clause) in its own line plays very nicely with version control systems.
Yeah, respecting literal line breaks would break some things, but simply replacing the two spaces at the end of a line with something visible (maybe two underscores if that doesn't break something else) should work.
Pandoc's extended markdown allows you to use a backslash at the end of the line to indicate a line break. (Think of this as a backslash-escaped newline, if you want.)
