I pretty much agree with you here, the torque wrench is at best a useful guideline for those who have not had the opportunity to work on 100s of bikes and develop a "feel" for what is right.

Today I use the torque wrench only on crank fixing bolts (square taper), mostly because of the unique nature of that interface, and also on modern stems with the face cap and 4 screws.

and maybe on BMW head bolts which are a discussion in themselves.

What I was trying to get to in the original post was that there's a phenomenon of "running torque" as the fastener if tightened down to approach the final torque value. On the really high quality cold forged crank arms (Shimano, TA Zephyr and Carmina) the torque value ends up being a very solid 260-280 in/lb. This is hard to describe, but it's where the increase in the running torque indicates that the fastener is sufficiently tight and the square taper joint is locked together. Exactly what you feel when you do it by hand. For classic Campy this ends up being about 220 in/lb. What I've experioenced on the Stronglight cranks (49, 57, 93) is that you never get to anywhere near that torque value. You just keep turning the wrench ! This may account for a large number of cracked crankarms.

For other types of assemblies in which failure of a fastener would be of high consequence (I'm talking aircraft and spacecraft here) the torque wrench provides a valuable "witness" that yes, in fact the bolt was tight when it left the shop. In many cases there is an actual step to accumulate evidence of this - do you know what a "Jo-Bolt" is ? Used in aerospace applications, the head breaks off when the specified torque is reached. Fifty bucks a shot.

But what do I know ?

/markp