No, the A knot has the down-the-cable runs overlapping the around-the-cable loops. Which means if one of the down-the-cable runs lets go, it loses tension on the next. In B, since the around-the-cable loops are overlapping the down-the-cable, if tension is lost at one end there is less tension lost at the next step. The friction in A when the end is loose is much less than the friction in B, due to the entry point in B being under pressure of the loop around the cable.
Aaaah, yea, that does make sense. The loop pinches both ends, so a loose end doesn't matter as much. Sorta requires the knot to be pressed against something, but that's probably frequent enough in practice / can be designed for (e.g. a knot that's bigger than the gap between things it's wrapping).
