There should be a speed test showing the issue (if any) between compiler optimization and January

[gerring]

It would be good to have a speed test to show where the current nD looping under performs against the compiler optimized basic loop. If January is slow then the test could be commented out until the team decides to fix something.