I was wondering about the same thing!
You’re absolutely right about the physics. According to Newton’s second law, the acceleration from a force is equal to that force divided by the mass of the object (a=F/m). Distribution of mass makes no difference.
Different mass distributions act only act differently when you’re applying torque* on your yoyo. The two ways you can do that are by throwing it and binding it (i.e. accelerating or decelerating the spin) or pulling the string at an angle so it rubs up against the sides a little bit (making the yoyo tilt a little bit).
My theory behind floatyness is that it’s actually an illusion. When you play with your yoyo, you get a feel for how you think it’s going act. Really high rim weights make the yoyo harder to torque, so it feels like it’s harder to make it move when you’re throwing it or pulling it at an angle. That makes the yoyo feel heavier than it really is in those situations. In situations when you’re not applying torque, the yoyo might feel mysteriously light and nimble by comparison. Therein, I believe is the illusion of floatyness.
*Forces create torque when they make things rotate. Rotating objects are really cool, so they get their own rules in physics. The torque created by a force is equal to that force times the distance between where the force is applied and the axis around which that force makes the object spin. The acceleration in the yoyo’s spin caused by a torque (it’s angular acceloration) is a function of how heavy different parts of the rotating object are and how far each of those parts is from the center of rotation i.e. the yoyo’s weight distribution.