Spawning a whole new python process and then re-loading all of Airflow
is expensive. All though this time fades to insignificance for long
running tasks, this delay gives a "bad" experience for new users when
they are just trying out Airflow for the first time.

For the LocalExecutor this cuts the "queued time" down from 1.5s to 0.1s
on average.

Data on this for a simple 10-task sequential DAG:

SELECT execution_date,
    min(start_date - queued_dttm) AS min_quued_delay,
    max(start_date - queued_dttm) AS max_queued_delay
FROM task_instance
WHERE dag_id = 'scenario1_case2_03_1' GROUP BY execution_date;

This was discovered in my general benchmarking and profiling of the scheduler for AIP-15, but it's not tied to any of that work. There are more of these kind of improvements coming, each unrelated but all add up.

^ Add meaningful description above

Read the Pull Request Guidelines for more information.
In case of fundamental code change, Airflow Improvement Proposal (AIP) is needed.
In case of a new dependency, check compliance with the ASF 3rd Party License Policy.
In case of backwards incompatible changes please leave a note in UPDATING.md.