People must be taking pictures of a bowl of fruit with these cameras, because any time I try to take a picture of a person who as much as blinks, the movement is enough to confuse the software into mega-blur mode.
That might be true but I find it not so important because I often hold down the button to take the picture so I get to take at least 20 snaps at a time and let Google's algorithm choose the best one and often use others for a collage. Call it machine learning or AI but the end product usually looks amazing.
Do you use this feature often?