Allow ookla download to accept a date range request and aggregate stats for all the downloaded stats

[butchtm]

This would be better if it accepted a date range - e.g.
start = pd.to_datetime('2019-01-01')
end = pd.to_datetime('2022-08-01')
speeds = ookla.get_ookla_data(start,end)

(Resulting dataframe would be a single aggregated dataframe of all the Ookla parquet files inside the date range, aggregated by mean(up/down), sum(total_devices) )

Originally posted by @AnthonyMockler in #127 (comment)