Following data (collection) idioms, there should be a task to generate this data into ~/cirq-results/ and then subsequent code can consume it from there. We should also upload a copy of it to a figshare dataset for fetching during tutorials / testing.

cc @ncrubin