This fellow's “editorial” is badly flawed.
His premise is that someone becomes mentally incompacitated, regains awareness, finds himself in a nonmoving car, then stomps on the accelerator. The writer suggests that since the car has creep mode enabled and the car is still, that it must be in a mode in which pressing the accelerator doesn't cause the car to move. He uses this to suggest that creep mode is safer.
Assuming his unlikely circumstance of sudden, total, transient mental incapacitation, if he had creep mode enabled and the car in drive or reverse, the car would have been moving under power during the mentally vacant episode. There is risk here that isn't addressed. Next, he assumes since the car isn't moving, the car must have been placed in park or neutral. If the car was moving under power while this driver was mentally vacant, there's a fairly good chance the car isn't moving because it has hit an obstruction, not necessarily because the car is in neutral or park.
To assume that the nonmoving creep enabled car had to have been placed in park or neutral while a similar nonmoving noncreep enabled car might not have been placed in park or neutral is the crux of his argument. There is nothing in the nonuse of creep mode that precludes this car having been put into park or neutral, so there's no advantage there.
To use this unlikely scenario to suggest the use of creep mode is safer than not using it makes no sense at all.
And now this silly argument is being quoted as a authoritative source for the increased safety of using creep mode. This isn't science it isn't even well thought out.
If you like creep mode, use it. If you don't like it, turn it off. But don't try to tell me it's safer to use creep mode based solely on this “editorial”.