I dabble in industrial reliability, Root Cause Analysis, etc. Not that I'm good at it, but even a blind squirrel finds and acorn sometimes. That said, I'm just baffled why Final drive failures are still an issue all these years later. It's almost expected, which is less surprising. Hopefully this won't degrade into a pissing-match-turned-name-calling session only to get punted downstairs, but I'll take that risk. My position is that these failures are not unusual, and that is interesting. So I did a little reading. I determined that a common, if not THE common failure is to the large single groove ball bearing, measuring 85x120x18mm or somesuch. Failures seem to be noticed after damage to seal occurs, but not always. In many cases, it seems as though the retainer damages this seal, but perhaps not always. I am suspicious of lubrication in both cases. Stay with me. I later determine that it is a commonly available (to industry) SKF or FAG bearing, but may be cross-referenced to any number of other makers. I also happen to have a copy of the SKF bearing installation and maintenance guide on my desk, so I flip it open to the lubrication section Page 87 HERE Get ready to check my math. Using a 102.5 mean bearing diameter, and 150/70-17 rear tire, I get 933rpm at 70mph, and 1197 rpm at 90mph, which according to this chart, I say means that you need a lubricant with a viscosity between 10 and 15 mm^2/s cSt. The next table gives ISO grade requirements at operating temp, I used 25-60c as a good range, although that may be a tad low. What do see is that an ISO 22 (SAE 75w is very close) or so is recommended. The chart assumes a VI of 95, and several of the gear oils I reviewed showed VI well in excess of this. My assessment at this point is that BMW designed a gearbox and specified a fluid appropriate for the gears but not for the bearings, especially when cold, intentionally or otherwise. My thought at this point is that a thinner gear oil with high pressure additives (moly?) would extend bearing life while not compromising gear life. Run-to-fail is a legitimate maintenance strategy, and perhaps the egg heads back in Germany decided that the bearings were cheaper than the gears, so f*$% em. Looks like a straight weight synthetic may serve the best in this application, still researching to see if that exists. Feel free to shoot holes in this. I assure you, nothing you can think of will offend me, and I'm all about learning. P.S. Still want a GS.