This change adds two types of tests:
1. First some regular Test* tests that functionally
ensure the test cases are correct.
2. A few Benchmark* tests that then allow us to
measure Otto's performance under various functional scenarios.
The basic tests are just sorting, but I'm planning over time
to bring in more advanced tests (like the CryptoAES) from
popular benchmarking suites (JetStream for instance.)
I intend to run profiler on these tests to find any
ottlenecks or low-hanging fruit that can help speed up
Otto for real-world use-cases.
Next up after this is to add some heavy underscorejs
benchmarks.
Responding to PR comments
