As you know the standard rubber bushing has an inner and outer metal sleeve with the rubber vulcanized to the sleeves. In the case of a basic control unit (wishbone), the outer sleeve is pressed into the wishbone, making that part rigid; when you tighten the bolt, ca 45 foot pounds, it's the inner sleeve that becomes rigid to the body part. The only flexing of the component is via the rubber itself, a few degrees in one direction, a few degrees in the other direction. Thus if you were to tighten the bolt with suspension parts hanging, when you lower the car, all the flex on the poor rubber is in one direction. On the other hand tightening (properly) at road level, which represents the average height of the car, this is the 0 stress position of the rubber. So as Jason pointed out, doing it the wrong way leads to premature failure of the rubber bushing (especially when you are reassembling the suspension without the bushing renewed). On a super light car like a Lotus, tightening in a jacked position will actually change the ride height of the car when you lower it, not to mention making the suspension quirky, as the rubber of the bushes themselves create unequal resistance to the suspension's movement.
I think that this thread shows, that no matter how experienced a mechanic you are, when working an a component that is unfamiliar to you, you need to carefully examine and catalog the parts as you remove them so you can reassemble them in the right order. There are lots of situations where shims are used to compensate for irregularities in the build; surprise surprise, not all 164s are the same. Something as simple as the AC compressor bracket is a case in point, where a shim might be added on one car at the factory to help align the serpentine belt, a shim here, a shim there, basta. Next car, please, oh, no shims needed, basta.
white '92 164 S (71K)