This ticket was triggered by the need of splitting 'unit-test' task to reduce Evergreen test runtime (WT-4410). In order to effectively group tests into multiple buckets (or Evergreen task) with similar amount of runtime, we need a script to run and evaluate the runtime of each test, calculate the expected runtime of each bucket based on given number of buckets, and fit tests into those buckets to achieve an evenly distributed result in terms of bucket/task runtime. 

We can periodically run the script (e.g. once a new Python test is added) to rebalance the tests across buckets/tasks. 

We should also consider making the script as generic as possible so that it could be reused to split tests for other Evergreen tasks.