Me too.  But for the truly anal retentive do-it-yourselfers out there,
tools are available to "do it right".  ;-)

Indeed- you're likely to get an error through the friction of the rod
turning in the shock and the bearing.

That's the problem.

I can usually get quite close to the correct setting by 'feeling' it. I
doubt it's a super critical setting. Haven't had any problem with changing
inserts on any car I've owned so I must be doing it about right.

Actually, you would not want to torque the strut rod.  You would want to
torque the nut and hold the strut.

Only problem being that a socket can't be used on the nut because the
hex recess in the strut rod comes up through it's middle.

There are short box end and open end extensions for torque wrenches that
allow this kind of operation.  The extension has a multiplication factor
to be applied to the reading on the torque wrench to get the actual
applied torque.  Or you could calculate the value yourself if you're sharp.

I checked my E39 - that has a hex (allen socket) in the top of the rod
which you use to stop it turning. To torque correctly you'll need one
which fits a torque wrench and a suitable spanner for the nut. Although
I'd guess there is a special tool. I'm pretty certain my long gone E28 had
the same arrangement - it's the only one I've changed the struts on.

Hello:)   I had this problem about 2 years ago on the passenger side and
took it in to the mechanic.  I remember him taking a large pair of
channel lock pliers to hold something under the fender while he
tightened the top strut nut.  My mechanic is now gone and I peaked under
the fender and dont see where could of put it to stop the strut from
spinning while tightening?  I dont want to put it on the chrome rod
unless that how its done? I put the car in the air and grabbed the tire
at 9 and 3 'clock and can see the whole assembly move.  Strut mounts
were new about 2 years ago.  I was hearing a noise over bumps on the
left side that why I went searching for the problem.  Thank :)Nate