This PR:

• Introduces a new CSE benchmark based on the "baby example" from Griewank & Walther (2008). This is a simple example function in two variables that requires careful selection of the operation order in the common subexpressions to represent the original expression in its simplest form.
• Introduces a new CSE benchmark based on the "lighthouse example" from Griewank & Walther (2008). This is a simple example matrix function in four variables where one of the two matrix entries is a common subexpression of the other.
• Slightly refactors the existing time_cse benchmark.

This PR introduces some test_ methods in addition to the time_ methods that will be run by asv. These are included here for bookkeeping with the hope that in the future SymPy's benchmarking will support a wider range of benchmark types (see this SymPy mailing list thread and this asv GitHub issue).