My first thought was that the app is going to be used "for real" for a first time at peak volume, but with a known number of users / data points to be gathered. Both of these requirements are unusual.

The first is detrimental - the inability to pilot real use, or "ramp up" gradually before the caucus makes it hard to avoid a "big bang" go-live and consequent issues surfacing at the worst possible time. As happened.

But the second is actually an advantage. Fixed and predictable live load rather than "we hope that it becomes very popular" give you a specific target to use in load tests ahead of time.

A senior engineer should be advocating for that load test at e.g. 150% of the real load.