on the complexity of Shor's factoring aglorithm with three quantum registers
It is well-known that the Shor’s factoring algorithm uses two quantum registers. Recently, by investigating the algorithm with three registers, the authors (http://doi.ieeecomputersociety.org/10.1109/ISISE.2009.86) argued the claim that the Shor’s algorithm runs in polynomial time is false, because Shor had wrongly treated a joint probability as a conditional probability. The argument is very short, and seems sound to me. If so, that’s big news.
Who can refute the argument of http://doi.ieeecomputersociety.org/10.1109/ISISE.2009.86