This is not at all a surprising development.
In fact similar "tinkering" was done to the A-DOT reports which were performed by Arizona State University that provided specious conclusions and suffered from a series of flaws including insufficient empirical data and instead contained heavily massaged and manipulated statistical analyses, also ignored key times of the day where accident rates occurred, and ultimately buried their preliminary findings that rates did not decrease (but were swapped in type with an increase in rear-end collisions).
And yet, despite these egregious flaws, Redflex and Scottsdale cherry picked the massaged results to misinform the public about how successful the program was.
Now Arizona is subject to a statewide deployment of this dangerous and misrepresented technology.
It is an Orwellian nightmare...