I don’t have my Mac solenoid in front of me, but assuming it matches the pictures I found online, the ports on the solenoid are labeled with numbers. Looking at the nameplate side of the solenoid, the ports are numbered 2, 3, 1 from left to right, with port 3 being under the nameplate. This translates into the following labeling in the HDi manual:
1: P
2: A
3: R
For your application (external WG), the solenoid will be connected exactly as shown on page 9 of the HDi manual. So, port 1(P) will connect to the turbo, and port 2(A) will be connected the upper port on the WG. Port 3(R) is left open. Note that the drawing in the manual could be misinterpreted and read as port 3 being the “P” port, since the arrow does somewhat look like it’s pointing at port 3 (the middle port). Perhaps that’s the source of your confusion, because you said that when you blew into the “P” port when the solenoid was de-energized, it was venting out a side port? If you are blowing into the actual P port (port 1) when de-energized, 1(P) is blocked with no airflow possible, and 2(A) to 3(R) is open. When energized, 3(R) is blocked and 2(A) to 1(P) is open.
In operation, this means that when the solenoid is energized, boost pressure passes through port 1(P) to port 2(A), applying boost pressure to the top side of the WG and preventing it from opening, which is how boost greater than the WG spring pressure is achieved. When de-energized, the upper port on the WG is vented to atmosphere from port 2(A) to port 3(R), allowing the WG to open. By rapidly shuttling between these two states, boost pressure above the WG spring can be controlled.
As a failsafe, since the de-energized state of the solenoid leaves the upper port on the WG vented to atmosphere (2(A) to 3(R)), the WG spring will then control the boost level, should the solenoid loose power for whatever reason.