It is more of a serious limitation of the concept, not a blunder per se. For the averaging to work, the risk per trade must be a constant value. This is in general not true or even applicable. Many traders adjust risk based on volatility. So yes, this analysis is flawed in general.
Someone...