Your text uses the parameter b<1, an effective energy sort of thing, whose shrinking leads to decreasing effective coupling, $$ \lambda ' = \lambda \left (1-\frac{3\lambda^2}{16\pi} \ln (1/b)\right ), \tag {12.28} $$ "infrared freedom", where the arrows in Fig (20.2) (b) indicate its shrinking.

But these are parametric slopes where b or the particular monotonic logarithm function of it have been eliminated in the ($m^2$, λ) plot provided. There is no sense of how fast a point in these RG trajectories flow on the plot, only the direction with decreasing b, towards the origin, absent a video of the flows. Any such monotonic flow function, eliminated, would give you the same trajectories.

Your text uses the parameter b<1, an effective energy sort of thing, whose shrinking leads to decreasing effective coupling, $$ \lambda ' = \lambda \left (1-\frac{3\lambda^2}{16\pi} \ln (1/b)\right ), \tag {12.28} $$ "infrared freedom", where the arrows in Fig (20.2) (b) indicate its shrinking.

But these are parametric slopes where b or the particular monotonic logarithm function of it have been eliminated in the ($m^2$, λ) plot provided. There is no sense of how fast a point in these RG trajectories flows on the plot, only the direction with decreasing b, towards the origin, absent a video of the flows. Any such monotonic flow function, eliminated, would give you the same trajectories.