Bill, Thanks for the link. Skimming it, I am wondering about the assumption that the orbits of hypothetical objects would be at their circular orbit velocities for their distances from the Sun. IF there is a "planet 9" on a highly elliptical orbit, wouldn't its velocity at perihelion be substantially more than circular velocity at that radius? If it happened to gravitationally collide with another object at perihelion, wouldn't it potentially transfer more velocity than calculated in this paper? And, then there is the potential for the smaller body to have been on a different highly elliptical orbit and having a velocity above its own V-circular at the time of collision.
Just spit-balling, it seems to me that they needed to use the escape velocity at a specific radius for the hypothetical collision in order to logically upper-bound the velocity that something could have by the time it reaches near Earth's orbit from a collision at the far edges of the solar system.
On the other hand, the probability of a real object experiencing an event near that upper limit seems tiny. So, if we see a substantial number of such objects zip by us, it would be very improbable that they came from such improbably gravitational interactions among bound objects, even if it is "possible".