Throttle Initialization-fetching

[Qqwy]

To make sure we don't open up our application for a DDoS, we should fetch data slowly, also during the initial initialization-fetch: Fetch in small batches of data, and then take a minute before fetching the next batch.

[InputUsername]

In addition, we should not start fetching new data until historical data is retrieved, otherwise EndpointCalls will end up in the wrong order.

[Qqwy]

@InputUsername That second notice is implemented in PR #176

[Qqwy]

This is nice-to-have, but currently the one thing that blocks adding Zeeguu to the project.

[InputUsername]

Implementing this would allow the Zeeguu dashboard to be added but would also break seeding; since the function to fetch historic data would instead fetch part of the historic data, and then be ran as a periodic task while fetching data. Seeding does not use periodic tasks and therefore cannot support the new function, and leaving in the old version of the function would be code duplication.