If you are defining a new user field function that uses components of a vector field function, your function will be clearer and execute a bit more quickly if you write the function directly in terms of the original vector function, rather than going through the derived scalar.
For example, suppose you want to define a new scalar function that is equal to the sum of the Cartesian velocity components. You could use the three intermediate scalar velocity component functions:
$Velocity_0 + $Velocity_1 + $Velocity_2
but it would be better to define it directly in terms of the velocity vector:
$$Velocity[0] + $$Velocity[1] + $$Velocity[2]