Thats all correct, except the design is based on the "required" wing loading for best performance at any given regime.
Think of it like this...
At stall, your wing is operating at maximum lift. Indirectly, this is at the maximum coefficient of lift. So, you are using all that wing can possibly create.
Now, give that wing extra speed and you no longer have to operate at maximum lift coefficient. So, in this case you could reduce the coefficient of lift of the airfoil and still have enough lift with less drag due to airfoil thickness or camber. OR...you could decrease wing area and keep the same airfoil resulting in less drag. Either one would be optimal at cruise, but wouldn't get the airplane off the ground at any less than cruise speed. So, it isnt optimal for takeoff.
Another way to think of it is that with higher speed, you can lift more with the given wing and it would be optimal to increase your "cargo" at cruise.
Another idea is, maybe the aircraft can give away a bit of slow speed performance to gain range or loiter performance. In this case the wing may be sized down to be more efficient at cruise to extend range or time aloft but will have a higher takeoff speed and required runway distance.
So, yes, once the wing is designed, wing loading is weight divided by wing area. But during initial sizing, you run scenarios of the optimal wing for each flight regime and the different wing loadings which are best for each. Then you select the lowest required and size the wing for that. If I had assumed stall is the determining factor (which I had, initially) I would have lost a measure of efficiency, which in this category, is very important due to limited weight and flight envelope.
Sent from my XT1585 using Tapatalk