Let's take a look at the early read from the Sturgis thing.
Sadly, there's no organized tracking of all the attendees. No master list to compare, no proactive testing, just scattered anecdote as individuals volunteer information. Is there some way we get a more direct read?
We could look at South Dakota stats, but that's only a tiny sliver because the vast majority of cases rode out of town and wouldn't be symptomatic until they were home for several days. (In a crowd over 10 days, it's reasonable to expect exponential spread, so the bulk of the infections likely happened in the last 3 days, August 14-16.)
Even though almost all of the cases left town, let's see what they left behind.
The event ended August 16. An infection in the last few days would start to show symptoms (if any) around the 19th to 23rd, and progressively worsen for a week. So most of the ones who seek a test will be getting tested in the last week of the month. Depending on test lag, they'd have results in the last week of August or first week of September.
Here's the chart for South Dakota through the 28th. That's a pretty remarkable signal...tripling the rate in just two weeks.
But South Dakota covers a lot of territory, and Meade County is only about 28k people, so are we really seeing a spike in Meade county? It's a pretty rural place, maybe they'r'e doing fine and one of the bigger towns is having an outbreak...
Here's a heat map of South Dakota. Sturgis is at the west end of Meade county, and it looks like they're now at one of the highest rates in the entire state.
But that doesn't prove anything, maybe they were already high before the event. Let's check the history...
Well, look at that. Almost no cases until mid August, then a few, then a huge spike. That's quite a jump in a county of 27,000. The lowest line above the origin is 200 cases/100k, they jumped from 40-ish to 240.
But Sturgis is nearly at the county line. What about the county next door with its 25,000 people?
And...boom. From 30 cases per 100k to over 160.
Remember that these are just the earliest confirmed cases, missing asymptomatics and many mild cases, and that the overwhelming majority of the infections rode away to other states.
This isn't a one-time event. The local people with infections have already spread it to their neighbors and coworkers. For every case we detect via testing we likely miss 4, and the downstream infections they cause will keep adding to the total for months to come.