I'm not asking the right questions.

Your previous survey only really included two kinds of interventions: changes to future consequences, and procedural changes. (You left out any type of cognitive intervention.)

Note that procedural changes can be immediately subject to "meta" akrasia, and a bad choice of consequence-changes can often increase akrasia in the short run, and in the long run are also subject to "meta" akrasia. (IOW, even if you invent an anti-akrasia pill, you still have to get yourself to take it -- which may not be as easy as it sounds.)

This doesn't mean that neither kind can work for some people some of the time, it's just that neither category will ever supply a long-term fix for everybody, all the time, even after you control for demographics and psychographics to constrain your definition of "everybody".