Under normal circumstances, I would agree with Steve about why some joints fail. Solder is pretty soft and cannot be relied upon for mechanical strength. In this case, I can't think what may have caused the pin to move other than thermal effects. I've seen dry joints on connector pins due to rough handling or vibration but, not being a metallurgist, it's hard to comment further.
When I was working in electronics it often seemed that many of the so-called electronic or electrical 'failures' were actually mechanical in nature or due to some outside factor such as dirt, water or heat. I guess the bottom line is that not much can really go wrong at the electron level in a well-designed system.
All parts are designed to work within reasonably safe limits for a certain length time. In the case of this part, the time elapsed last Friday night. But wouldn't it be gratifying to know whether the person who assembled this circuit board ever gave a thought to how many people would mess their pants in the fast lane when the fuel pump stops doing what it does best. The mind boggles... Hi, is that the MB service desk? I'd like to arrange an appointment for my wife and kids to come and meet assembler XY89.9 who was kind enough to leave his or her employee ID in my fuel pump relay.
1990 230CE 150k+ UK spec.