1. The best discussion of the question of the rate of false convictions of which I am aware can be found at the excellent Texas jurisprudence blog gritsforbreakfast.blogspot.com. From memory, the blog's host estimated it at 1.5-3%. Each nation's mileage may vary, but 1.5% doesn't qualify as vanishingly small, IMO.
2. You should ask the person who brought it up. I surmise that it was brought up to illustrate that a guilty verdict is not certain proof of actual guilt. You seem to hold a view that is very close to this, based on your comment at 9:42 AM.
So we can be at least 97% sure that the court correctly convicted her - good enough for me.