The mechanic could be correct. In either the case of the AAV or the vacuum leak, I believe the scenario is the engine is getting too much air which causes the engine to race. An AAV that is stuck open is similar to a vacuum leak or the idle screw being wide open.
I think on the d-jet the AAV gets it's air from a tube in the front if the throttle body above the throttle plate. If you partially block the off, does the swinging idle come under control? BE CAREFUL though. The suction on that tube is very strong. I think I've gotten my finger caught on it (and stalled the engine) before.
Fonzi is right. A sticking AAV is essentially a vacuum leak. There are several potential leak points:
- hose from back of intake manifold to Manifold Pressure Sensor
- connections at rear of intake manifold, passenger side that go to door lock system and transmission.
- small plastic line that runs from throttle body to a valve on passenger side fender and on to distributor (controls distributor retard)
- short hoses that connect upper and lower halves of intake manifold internally (internal, can't be seen)
- hoses on either side of AAV.
- any others Fonzi?
Regarding blocking the port on throttle body. That port supplies idle air to both the AAV and the idle control valve. If you block it, engine should stall if there are no vacuum leaks because that would block total idle air supply when throttle plate is closed.
Generally, if you have serious vacuum leaks you cannot adjust idle speed using the idle screw. That is why I suggested getting car well warmed up, then set idle speed to 700rpm. If your idle screw has no affect, then start looking for vacuum leaks (AAV not closing or elsewhere)
I know I was advised many years ago by an MB mechanic to search for vacuum leaks to solve a similar surging problem. Even to use a smoke machine to find the leaks. But it turned out to be a sticky AAV and in time I found that many others had same problem.