Standardize how our fakes are used so that a test case can use a
simpler mechanism for providing large, complex data sets, as well
as represent queries over time.