Execute parallel tasks for nested tasks in a job spec

[pandemicsyn]

Given a yaml such as:

jobs:
- name: g-to-p-job
  tasks:
  - - tap-gitlab target-postgres
     - tap-github target-postgres
  - dbt-postgres:run

Today we create one dag with two connected tasks:

#Dag -> task1 -> task2

task1 = meltano run tap-gitlab target-postgres tap-github target-postgres
task2 (depends on task1) = meltano run dbt-postgres:run

But it should be "trivial" to parallelize the subtasks rather than flattening them into a single task:

#Dag -> task1 (parrallel subtask1, parallel subtask2) -> task2

task1 = (subtask_1, subtask_2)
    subtask_1 = meltano run tap-gitlab target-postgres
    subtask_2 = meltano run tap-github target-postgres
task2 (depends on task1) = meltano run dbt-postgres:run

To do so, we'd need just trigger subtask creation at this check:

files-airflow/bundle/orchestrate/dags/meltano.py

Lines 145 to 147 in ba72d83

	if isinstance(task, Iterable) and not isinstance(task, str):
	run_args = " ".join(task)
	else:

This parallel execution isn't native to meltano and wouldn't be support yet by meltano run natively, but it would at least allow airflow users to leverage parallel subtasks.

[pandemicsyn]

@aaronsteers @tayloramurphy might be nice to try to squeeze into a future iteration. This would be a pretty simple change (weight 2) but should teamed up with #17 (adding some test coverage).