Periodic rebalancing already brings additional return. Threshold-based rebalancing is only
marginally better.
You can check the
research report (page 57) cited in the article. For convenience, I reproduced the data below:
Return benefits of rebalancing once a year (250 market days):
0.15%
Return benefits of using 20% band:
0.16-0.20%
Advantage over annual rebalancing:
0.01-0.05%
I think opportunistic rebalancing is not very attractive for DCA-investors. There is added complexity for the small gains.
This method is definitely more suitable for non-DCA, opportunistic/speculative portfolios. These investors like to time the market, and this method works well for them as it provides a mechanical way to do so.