"onthedials" may have hit the nail on the head.
The WOLNH waypoint is coded in a GPS as a 'flyby waypoint and not a 'flyover' waypoint. When the gps senses that a waypoint is approaching, it calculates a distance to turn anticipation. When the tracks are the same but in opposite directions the distance to turn anticipation is infinite. The out bound track to WOLNH will be very close to the inbound track back to YWOL and so the distance to turn anticipation will be quite large.
So the scenario is this.....
You get to the missed approach point and decide to continue missed approach. You then press 'GOTO' WOLNH. The system sends you off in the direction of WOLNH, but also calculates a turn at WOLNH to go back to YWOL. Lets say distance from missed approach point to WOLNH is about 6nm. Turn anticipation for a reciprocal course may be about 4-5 miles. You are still tracking to WOLNH, then about 4-5 to run to WOLNH you pass the turn anticipation point and because inbound and outbound tracks are so close together you also pass the midpoint of the turn. This forces to system to assume you made the turn, and 'flewby' the WOLNH waypoint, thus sending you back to YWOL.
If you remove YWOL as the last flightplan waypoint, or pick some other waypoint....you shouldn't have a problem.
This is the problem with different GPS receivers. The GPS coding specification doesn't define what to do if the turn at a waypoint is greater than 120degrees. So each manufacturer can code their recievers differently. It would appear BENDIX/KING recievers always calculate a turn anticipation distance. I think in the same scenario GARMIN recievers would make you fly a course reversal at WOLNH, thus eliminating this problem. I have flown this approach with a GARMIN 430/530 and not had any problems.
Alpha