So you feel that randomly applying corrections that you make up to data from specific intervals (between date X and Y) is totally kosher in terms of modern scientific practice? If the data doesn't fit, you don't make it fit.