MathOverflow will be down for maintenance for approximately 3 hours, starting Monday evening (06/24/2013) at approximately 9:00 PM Eastern time (UTC-4).

It seems to me that Baker-Cambell-Hausdorff is just (noncommutative) algebra. A nice closed formula formula was found by Dynkin On a representation of the series $\log(e^x e^y)$ in non-commuting $x$ and $y$ via their commutators Math. Sb. 25 (67) 1949. The book "Methods of Noncommutative Analysis" by Nazaikinskii, Shatalov, and Sternin has a nice presentation.
There is also an interesting paper of Mosolova: New formula for $\log(e^B e^A)$ in terms of commutators of $A$ and $B$. Math. Notes. 29 1978.
It seems to me that Baker-Cambell-Hausdorff is just (noncommutative) algebra. A nice closed formula formula was found by Dynkin On a representation of the series $\log(e^x e^y)$ in non-commuting $x$ and $y$ via their commutators Math. Sb. 25 (67) 1949. The book "Methods of Noncommutative Analysis" by Nazaikinskii, Shatalov, and Sternin has a nice presentation.
There is also an interesting paper of Mosolova: New formula for $\log(e^B e^A)$ in terms of commutators of $A$ and $B$. Math. Notes. 29 1978.