Personally, I get very little use out of this technique, since my problem tends to be uncertainty about the likely consequences of my actions, not uncertainty about which outcome would be best.

Have you tried it on a micro-scale? I employ a modified version of this technique as a constant motivation tool. (Eg, I don't feel like going to the gym and prefer to read things on the internet, so I query my future self from 4 hours ahead and future self from a couple days ahead for each Everett branch of action and poll my imagination of their opinions. Invariably the 4 possible-future me's all outvote present me and force me to the gym.)

I find it very good for peer pressuring myself with my future selves, but it only works on things I cognitively know the 'right' answer to yet am emotionally unconvinced by. It also helps exceptionally well for hyperbolic discounting. I think that Harry is using a similar tool to line up his emotions and motivations with what he knows cognitively and to avoid the shortsighted path (Kill 2/3rds of the Wizengamot) in lieu of the path he'd previously decided on.