Since the 80s, academic papers document the "progress" of myoelectric control error rates. These are usually determined in safe, constrained laboratory settings. These error rates are nowhere even near what would be needed for a safe, and reliable, everyday function. They have not improved at all over a period of forty years. This seriously questions oversight, focus, goal finding, and overall respect towards the amputees that rely on this technology. Having a prosthesis that "always" drops items is expensive: you have to replace broken cups or plates or glasses. Calculation examples and comparison to error rates of body powered arms in the referenced blog article.