My take on the definition a planet

As you know (and if you didn't, read the comments on the post below describing Pluto), but there is a legitimate debate whether or not the definition of a planet should be dynamical (orbital parameters, mass, etc.) or geophysical (essentially, is it spherical and orbit a star?) in origin. Personally, I prefer the dynamical view, and find the geophysical view too inclusive. Sphericity means its mass is above a certain value, or that the mergers that produced the object were at sufficient velocity to melt the product such that is cooled into a spherical shape. The number of spherical objects in the Solar System is large, and as a result personally don't feel that should be the primary definition of what makes a planet. It doesn't bother me that Ceres is not taught in schools because, frankly, it isn't that special. It is simply the largest asteroid in the asteroid belt - it is not even the only spherical asteroid.

I prefer the dynamical view because it better incorporates our current understanding of star and planet formation. A cloud of gas and dust gravitationally collapses into a star. As it does so, a thin disk of gas and dust form around the star. Inside this disk, collision between dust particles result in proto-planets, which then coalesce through collisions into planets. These objects are sufficiently massive that, through gravitational interaction, they most "clear" their surroundings of gas and dust - opening a gap in the disk (which has been observed in several systems). The friction inside this disk is believed to lead to circular orbits, but this doesn't have to be the case - as recent observations of exoplanets indicate (as you point out). This does mean that the objects orbit in the same plane.

If one was to define a planet based on this model for planet formation - which may or may not be the best thing to do - the requirements would be:

(a) "By far" the most massive object in its vicinity, a signature of the object clearing out a gap in the proto-planetary disk in its formation. By this criterion, Ceres would not be a planet because it is located inside the asteroid belt, whose total mass is larger than that of Ceres. The same is true for Pluto. The mass of known objects in orbits similar to that of Pluto is greater than that of Pluto. Note this criterion does not demand sphericity, just that the object is much more massive than the sum of all objects in similar orbits.

(b) The object orbits the same plane as other planets in this system. Now, this is a little tricky formally because it is circular - an object is a planet because it has an orbit like other planets. However, in our Solar System, the orbital plane of Mercury through Neptune varies by only a few degrees. [This is true for objects in the asteroid belt as well, which is the asteroid belt is often called a "failed planet" - it would have merged into a planet if not for its location near Jupiter]. Pluto's orbit, and that of the other massive spherical KBOs, is far off the ecliptic.

This definition is far from perfect, but I personally feel it better represents our current understanding of planet formation. Admittedly this is not the official IAU definition, but the official IAU definition incorporates these two elements. This definition is also general enough that it incorporates many of the "odder" exoplanets you mention - the ones with high eccentricity, the giant planets very close to their star, the planet that orbits its star backwards. As more exoplanets are detected, and these planetary systems and our own Solar System are better studied, the definition will and should change again. The new IAU definition I believe is a step in the right direction, and Dr. Stern's is a step in the wrong one - I feel that too many objects satisfy this criteria for this designation to have any meaning. This view is shared by a vast majority of astronomers I know (including the ones who study exoplanets and Solar System objects), and I believe that of a majority of astronomers - though far from unanimous as you point out.

And yes, dwarf planet is not a great name. Plutoid is better. I do feel that looking for self-consistency in astronomy nomenclature is futile, because frankly there is none.

Lastly, it is not clear that Pluto satisfies the dynamical (Dr. Stern's - quoted in the comments below) definition of a planet being "any non-self-luminous spheroidal body in orbit around a star." The center of mass of the Pluto - Charon system is outside of Pluto. Also, the gas giant planets radiate more light then they receive from the Sun. Therefore, they could be considered self-luminous, and would also not be considered planets under this definition.

[For those reading the comments below - Hydrostatic equilibrium is when, everywhere inside an object, gravity is balanced by an outward force. This is true for all stable objects (essentially by definition) - spherical or not.]

What do you think? What is your opinion? Does this make sense? Let me know, I'm genuinely curious.