[Mark McLeod]

Registration is up about 3% over last year at this point in registration.

Here are two three revised registration progression plots of the delayed opening.
The first just jumps (pretty much a straight line) from the morning after the late registration opening to corresponding surge markers that are present every year.
The second is the more accurate one that simply slides the opening day as Jon suggested earlier. Just plotting the actual delayed date looks better, but makes it harder to see the correlation between that day-early District arc and the 2012 regular arc.
The third uses the real dates for registration who's start dates spread over a couple of weeks in the past decade. Mile-marker events such as 2nd District opening and unrestricted Regionals, even US holidays are uncorrelated giving subsequent years a slanted look. It's most noticeable between the light green 2016 line (earliest registration ever opened) and the white 2017 plot (latest it ever opened).

The first plot variation better compares the fast fingers on the first day, while still correlating with the standard markers (where the surges are every year), such as the 2nd District event registration day where a lot of MI teams come out of hiding.

See if this is understandable and give me your preference.
I played with stretching the interpolation, but that didn't look much different than a straight line.