How does the orbital motion of reduced mass tell us how the individual planets/stars move?

I think the choice of a specific set of coordinates is somewhat blurring the answer to your question.

Let start from the beginning and let us follow the main steps required to arrive the Kepler's first law in the form of the equation (29) in your figure (the polar representation of a conic).

The usual strategy is to write down the equation of motions for two point-like masses $m_1$ and $m_2$ in an inertial reference frame:

$$ begin{align} m_1 {bf ddot r}_1 &= Gfrac{m_1 m_2}{|{bf r}_2-{bf r}_1|^3}({bf r}_2-{bf r}_1) m_1 {bf ddot r}_2 &= Gfrac{m_1 m_2}{|{bf r}_2-{bf r}_1|^3}({bf r}_1-{bf r}_2) end{align} $$

It turns down that in terms of the following new variables the equations are simpler: $$ begin{align} {bf r} &= {bf r}_2-{bf r}_1 {bf R} &= frac{m_1 {bf r}_1 + m_2 {bf r}_2 }{m_1+m_2} end{align} $$ Indeed, we get the following equations: $$ begin{align} {bf ddot R} &= 0 tag{1} mu {bf ddot r} &= -Gfrac{m_1 m_2}{|{bf r}|^3}{bf r} tag{2} end{align} $$ where the reduced mass $mu$ is defined as $frac{1}{mu}=frac{1}{m_1}+frac{1}{m_2}$. The interpretation of equation ($1$) is that the center of mass moves with a uniformly rectilinear motion. Equation ($2$) can be interpreted as the equation of motion of a body of mass $mu$ subject to the force due to Newton's gravitational law. It is the orbit obtained by solving equation ($2$) that is shown in equation ($29$) of your text.

The inverse change of coordinates: $$ begin{align} {bf r}_1 &= {bf R}-frac{m_2}{m_1+m_2}{bf r} tag{3} {bf r}_2 &= {bf R}+frac{m_1}{m_1+m_2}{bf r} tag{4} end{align} $$ is the key ingredient to understanding how the trajectory obtained by solving equation ($2$) for the relative position vector ${bf r}$ can be translated in terms of the two bodies' trajectories in the original inertial frame.

Indeed, equations ($3$) and ($4$) directly show that at each time $t$ the center of mass $bf R$ and the two positions ${bf r}_1$ and ${bf r}_2$ keep aligned. Moreover, the distance from the center of mass is proportional to the ratio between each body's mass and the total mass.

Therefore, if, by solving the equation ($2$), we obtain an ellipse described by the vector ${bf r}$, such that the origin of the vector is in one focus, equations ($3$) and ($4$) show that each body $i$ has a trajectory that is an ellipse rescaled by the factor $frac{m_i}{m_1+m_2}$, relatively to that described by ${bf r}$, but now the focus is occupied by the center of mass position.

Notice that the introduction of the relative position ${bf r}$ is something more than a purely mathematical passage. It is consistent with a description of the problem in a non-inertial reference frame centered on one of the two bodies (body $1$, with the definition above). Interestingly, the appearance of the reduced mass can be obtained as the effect of writing down the equation of motion for body $2$ in a non-inertial reference frame by taking into account the acceleration of body $1$ due to the gravitational force due to body $2$. In this way, we can see more clearly the coexistence of two different descriptions of the same phenomenon:

1. in an inertial reference frame, we have two bodies performing each an elliptic motion with the center of mass at the common focus.
2. in the non-inertial reference frame centered on each of the two bodies, the other is describing an elliptic orbit having the one focus at the origin.

A final word of caution is on the geometric description of the orbits. While in the case of the reference frames centered on one of the two bodies, there is a unique description of the orbit as a conic section having the other body at one focus, in the case of the description in an inertial reference frame, it is only in the reference frames where the center of mass is at rest that one can speak about conic sections. In the most general inertial reference frames, the orbit's shape is the composition of a pure translation with the conical trajectory.