Well this was not a trick. The battery timings that you can see after doing this 'so called trick' is based on some calculations. Now see, if you know that the estimates are based on past results ok. Now if you are going to estimate an amount which is based on few samples you will probably get an inaccurate result. But if my estimates are based on several samples then the results will be more accurate. Same is the case with battery, the more past results you will have the more accurate will be your estimated results.
I can prove you mathematically
I have two type of samples one is based on 1 day and other is based on 4 days data.
Let's suppose The data of 1 day only: 5% dropped in 2 hours. (data 1)
The data of 4 days:
1st day= 5% dropped in 3hours
2nd day= 5% dropped in 1 hour
3rd day = 5% dropped in 0.5 hour
4th day = 5% dropped in 2 hours
It means on average 5% dropped in (3+1+0.5+2)/4 = 1.625 hours (data 2)
Hence proved that if 5% dropped in 1.65 hrs (as per data 2 above)then the remaining 95% will be dropped in (1.65/5x95) 31 hours which is 1 day and 7 hours.
Whereas if the 5% dropped in 2 hours (as per data 1 above), the remaining 95% will be dropped in (2hrs/5x95)=38 hours which is 1 day and 14hours
You can see that the one which is based on 4 days data has estimated remaining time of 1day and 7hours which is less and hence more accurate than the one which is based on 1 day.
So clearing the data of extended battery controller from Apps will actually remove the previous data of your battery usage and the estimated time is now based on few hours which is inaccurate. Sorry for the so long explanation.
Genuine Method of Battery Saving: CLICK HERE