The minimum rate of descent speed (or Vimp) is slightly lower than the minimum angle of descent speed (or Vimd) for the same physical reason that Vy is slightly higher than Vx.
In both cases, we compare a rate of altitude change per horizontal distance (in the air mass) with a rate of altitude change per time elapsed.
They relate to each other through the airspeed (and cosinus of the angle of descent which can be ignored on first approach).
dy/dx = v * dy/dt * cos(angle)
Now it is easily shown that a somewhat "simple" function that goes by 1 minimum or by 1 maximum, gets its minimum moved earlier on the curve
or its maximum moved later on the curve when the function is multiplied by the variable it acts upon.
I can send the math demonstration by MP, if you want.
So, if one names Yx(v) the function of angle of gliding descent per the airspeed, and Yt(v) the rate of gliding descent per the airspeed :
Yt(v) = v * Yxt(v)
(ignoring the cosinus of the angle)
and if Vimd is the minimum for Yx(v)
and Vimp is the minimum for Yt(v)
then Vimp < Vimd
Flying at Vimp, a lower speed than Vimd, one will fly longer although on a shorter distance
For Vx and Vy, we are looking for a maximum rate of climb,
Vx is the maximum for Yx(v)
and Vy is the maximum for Yt(v)
therefore Vy > Vx.
Flying at Vy, a higher speed than Vx, one will climb quicker although on a longer distance
Hope it clarifies the relationship between optimal angles of climb/descent and optimal rate of climb/descent.
Regards
Luc