A couple of weeks ago I tested fuel economy in my Volvo over a 5 mile stretch of freeway without, then with an ounce of acetone added to 10 gallons of gas. I found an immediate 6.5% drop in mileage between those first two tests. Then after 300 more miles I tested on the same tank of acetone-laden gasoline using the same conditions, and found I got an added 4.2% to mileage. I was intrigued.
Now I've been through 15 more gallons, and haven't added any more acetone. I did two runs tonight to check where the mileage was. Same weight in the car, same temperature outside, etc. Just about the only thing different was the song playing on the radio! Mileage was slightly better still. The first run I averaged 38.9 mpg, which is 5.5% better mileage than my original baseline, and the second I got 39.2, which is 6.3% better. Looks like I've got at least a +/- 0.4% margin of error.
So bottom line after all this testing is that the people who see better mileage will probably continue to see that better mileage once they stop using acetone. I still think I started with gunk in my fuel system that got cleaned out from the acetone. Mine is a higher-mileage car, 122,000 miles. Easy for me to believe that this is what's happening here.