Nope, but git has overall superior tooling, support, and community acceptance. It is generally a superior choice to non-distributed solutions and in absence of a verification of the patch commutability theory, I don't see a significant advantage to the way Darcs approaches the problem. The tree based approach is a little less flexible, but it's still very good.

git's only "theory" that could potentially be verified is merge resolution. which has been.

I think darcs approach is a little different.

git has plenty of properties that are supposed to be true. The developers write tests for them. The properties being tested could be proven instead.