You don't seem to be using git the way I do. Let me give a little example:
1. Download some non-versioned source from a website, untar it into dir
2. cd dir; git init; git add .; git commit -m 'initial'
3. work, hack, commit, etc
4. decide that I don't want the history and just rm -rf .git (or the whole dir)
The point is that you don't have two locations for one tiny thing (a repo and a working copy). Also it
works in-place, whereas you have to create the repo with SVN, then checkout the _empty_ repo into
another dir, and then manually copy your files there to add and commit them. Still sounds equally
By the way, if you just want to keep track of very simple changes to a pristine codebase, you can
omit the "commit" in step 2. and just diff against the index. Another tiny thing: Your example
doesn't work (for me), since SVN is too stupid to recognize files without file:/// prepended.