I would start by checking your eddy boards and make sure they are adjusted correctly. I just had to adjust mine and I had a problem that sounded like what you are going through.

I replaced my center ramp target an had the problem of the clock advancing when it was hit because I soldered the wire on the wrong side of the diod. Thay are also inline with the trunk hit on the switch matrix column

Go into feature adjustments and see if #14 trunk motor is set to yes to allow the motor to run
It may be disabled.
You can download the manual here:
http://www.ipdb.org/machine.cgi?id=2845

^ I hear ya. I'm new with this stuff too, and I have trouble correcting just about every issue that comes up. My TOM pop bumpers and slingshots work in test mode but not during play. I've cleaned and adjusted the switches and still nothing. It gets frustrating.

Any change when you put the original trunk eddy board back in? The game should register 3 hits on the trunk, then hopefully give you the option to start an illusion by hitting it into the trunk as you know. On mine, if you hit the trunk three times and then immediately hit the trunk while it's moving into position, it automatically starts the illusion. Maybe there's a connection there. Is your eddy board overly sensitive or registering double hits? If it was and if the double hit came soon enough, it might not move the trunk but just start the illusion instead.

When it starts the illusion does the trunk move to the varied appropriate position for each illusion?

Particularly with a fussy trunk...

I agree. If there's no better way to test this out, you could put the game in trunk test mode, and lift the playfield to see if the LED on the opto board is triggering at any point during the test. It should not be.

Just curious if the opto interruptor is in good shape (black disc on bottom of circuit board underneath the trunk). Any cracks?

Tiger saw does advance clock, but only after you lock a ball and hit it as a skill shot. Are you getting any credit dots on the display? I would put it in to switch edge test and see if you can create any phantom hits on the trunk.

Does this mean everything is working now?

This is a switch matrix problem, I believe. I don't think it's Eddy-specific.

Of course, you could disconnect the eddy boards one at a time to test my hypothesis.

Is there a fuse associated with the slings and pop bumpers?

I'm sure I have had the same when illusions start immediately after hitting the trunk 3 times, but it's not doing it now. I did play around with the adjustment menu A2 .19 which sets how many hits to start an illusion. It sounds to me like an over sensitive eddy sensor. some times mine registers 2 hits if the ball hangs around the area with a bit of spin on it.