Mike
Nice drawings, you put a lot of work into them. Your surmise about the pivot points is correct, the A end draws up first although the action is simultaneous for all practical purposes. The reason for the spring is still a question though. On all of the K brakes I have worked on the cylinder rod is not actually connected to the brake piston, therefore the piston pushes the rod out but does not retract it. The motion of the wheels causes the shoes to fall away from the wheels. With the spring positioned as you have it, the only thing it would accomplish would be to provide a spring action between the brake wheel and the brake cylinder lever. There could be an advantage to this however. Often, when setting hand brakes, the chain comes up tight just before the next tooth on the ratchet. When you back off to the previous tooth the chain has a little slack. A spring in the location pictured could provide just enough give to allow the chain to remain tight. However, I have never seen this on any of the D&RGW, WP&YR, or EBT cars we have.
Just another thought.
Jim Grigsby'
SVRR Lead Car man