Yes, I saw similar ideas ~ a year ago, & figured it was possible. It was just a matter of using sqlite's (excellent) sqldiff utility and wrapping a custom git diff driver. Took about an evening to get the proof-of-concept working and a little more fiddling for ease of use.

I don't have stress test results, but it should be similar to git. I think I remember getting it up to several hundred megabytes at one point and it was fine. I mostly use it for smaller sets of highly relational data that I want to track like I would source code.

By leveraging git & sqlite it lets me avoid writing a network sync implementation, architecture specific code, or patching any C code to recompile.