When you have an object per method you are really doing structured programming. When you have one object with three methods you are doing object oriented programming. Most people consider OO the better approach of the two.
I have done both (which shows my age) and prefer 00. The big advantages with OO are inheritance and data encasulation. The best use of structured programming that I found was the jods where millions of rows of data were being crunched in some nasty algoritm and run time was the major consideration. By doing some really bad programming, you could make the program run much faster.