Over the years several forum replies about seal replacements to tighten down the nut to the same exact position (paint marked) first before removing the nut. Pinion preload from new was set up properly by factory but now with worn bearings to follow those instructions your pinion preload will be much lower, possible to the point of almost feeling play in the pinion. This play gets bigger under loaded power conditions changing the critical gear face alignment plus causing the seal to not do its job if that loose.
Yes a shim pack behind the pinion gear I recall your correct after converting the Roto Crap to a real steel U joint driveshaft back in 2000. I have looked under several LR's now having 100 to 200K miles that the owner replaced the seal and nut to the exact position then having a seal leak again a couple thousand miles later. First checking the pinion to see if it has any loose play, slack in the bearings, yup had seen this on several at these higher mileage LR's, about half of them. Just think what that does to the R & P gear faces, gear whine and failures all prevented if set up properly. Preload torque readings in in/lbs up to ft/lbs to turn pinion whatever the specs dictates. This should be without engagement of pinion turning the ring adding a higher false reading. You can get a rather close idea between the R & P backlash readings especially if the R & P gears were dry not coated with oil cushioning the backlash clearance.
On another note there are two pinion preload torque reading numbers to address. One for new pinion bearings being higher allowing the bearings to break in and the used bearing preload torque numbers used with old bearings turning the pinion. Must add to do it properly pinion rotational torque readings should be done without ring gear as well without the oil seal adding increased false torque readings.
Good thing LR didn't use a crush sleeve for the proper preload, that's another POS "El Cheapo" production process the bean counters approved.
Burning rubber in reverse with a crush sleeve setup, pinion now wants to pull away from the yoke vs to the rear vs forward loading up the taper bearing, crush can become more crushed hence a loose pinion condition. Not a problem with a shimmed pinion setup.
Even brand new LR's at the dealerships back in 2000 i'd crawl under them on the showroom floor looking under plus noted a large amount in degrees pinion rotation before hitting resistance then rotating in the opposite direction until resistance felt. Had to take them out of Park first pissing off a few sales idiots. Way more like 3-4 times more degrees rotation than the 922K mile Dyna 60 diff in my F250. What's up with that I ask?
Sloppy drive train in an automatic who cares, in the 5 spd D1 it's noticeable and a PITA clunk clunk between accel to decel. Not the trans output spline to transfer gear spline failing syndrome. English go figure nothing surprises me anymore at 20 years of ownership. it owns me. Still love the big lump. .....~~=o&o>......

 

So basically just install the ujoint flange and torque to spec.

 

fishEH, you are correct but must add more to this topic.
Pinion yoke nut should be torqued to 130 Nm or 95.88 ft/lbs period. The pinion preload with used bearings should have 1.5 Nm or 13.28 in/lbs torque with old bearings, 3.0 Nm or 26.55 in/lbs torque when new bearings are installed..I rather deal in the in/lbs scale vs a low 2.212 ft/lbs numbers not being that accurate.
Pinion preload with yoke nut torqued down is corrected by adding in this case more preload as the used beatings have broken in and are worn hence adding a shim behind the pinion cone bearing at the yoke end of the pinion. Not after the cone bearing has been installed or before the yoke......~~=o&o>.......