It's a fallacious argument to use a specific date for the "reopening," as it's an ongoing event and a moving target. The act of reopening didn't do anything to the number, but the change in peoples' behavior brought about by the reopening did and will continue to do so. Beyond just the standard 1-2 week lag, I'd argue it took at least that long on top of that lag for enough people to "recalibrate" from sheltering in place to participating in the public sphere to the extent needed to be reflected in the numbers.

Part of the problem is people are trying to derive a clear signal from a very noisy data set. [Not directed at you] Be wary of anyone trying to make any absolute argument about the data that ignores that reality.