I didn't say anything about doubling returns. Perhaps you read the following sentence and thought it was referring to returns...
"The Daryanani study showed that an Opportunistic Rebalancing strategy more than doubled the calendar rebalancing benefits over a wide range of market conditions."
Daryanani stated this in the Executive Summary of his paper.
Regarding tax sheltered accounts, I agree these types of accounts would be best. However not all investors are U.S.-based and there are countries that don't have the long-term/short-term tax rates like the U.S. has.
In addition, Opportunistic Rebalancing can make sense even for U.S.-based investors using taxable accounts.
First, if you've been buying shares over time, some of these shares could have been purchased more than a year back. These are the ones you can sell when you rebalance.
Secondly, Daryanani talks about tax costs under the section, "Costs of Rebalancing" in his paper. So in some situations it makes complete sense to rebalance this way in a taxable account.
The main reason I like this method is because it's based on price rather than dates. And since price is really what creates a portfolio's imbalance, it's logical to use it as the prime determiner on when to trigger a rebalancing event.