I agree.

My guess is that it didn't though, or if it did, that it didn't work as advertised.

If a GPS had been involved and working right, the position should have been reported to within 10s of meters of his actual location - assuming that he was in a spot where the unit could have gotten a GPS fix.

On one of the PLB test reports I read it seemed like a number of the units failed to provide good GPS coordinates. This info about some GPS equipped units not sending good GPS position info was also reported on some Cospas/Sarsat site documentation I read.