We've gone from debating about whether or not there was a driver, to debating about whether or not Autopilot was turned on, to debating about what Autopilot even IS! To me, given what was posted on a prior page from the Tesla website, TACC is a subset of Autopilot so TACC in and of itself is not Autopilot.
But I don't think any of these semantic arguments are going to matter in the end. What they will be looking at will have nothing to do with whether TACC can or cannot be called Autopilot. They'll likely be looking at whether or not some features of the car encouraged the owner to perform dangerous actions. As such, they may look at the possibility the owner thought the car was steering itself when it was not, whether or not to make a recommendation that the action that engages autosteer should be separate from the action that engages TACC (rather than AP being essentially performing the TACC action twice), whether there should be more clear driver indications about when the car is actually autosteering, etc.
This will likely be done in conjunction with (but not depending on) the actual crash scene investigation regarding what the driver did and how much fault lies with the driver versus the car.
Mike