In science, classification should have some predictive value. You should be able to predict the characteristics or behaviour of a class member based on its classification. For example if you classify something as a fruit you could predict that its seeds should be on the inside.
Classifying Pluto as a planet is problematic because it lacks the behaviour and characteristics of something that formed according to theories of how planets form. If you tried to describe what a planet should look like based on the nebula hypothesis, you would not be describing Pluto.
Instead it has behaviour and characteristics a lot like the many trans-Neptunian objects, so it makes more sense to classify it that way.
