Yes all the talk about stretch is good theory. Stretch is the
most accurate way to measure tension. The torque required is derived by the bolt manufacturer, based on material properties, area of the section, the coefficient of fiction between the fasteners, etc.
I was on a job a few years ago where elongation and tension were the only criteria specified for the bolts (bolts were through bolts for cable clamps on a suspension bridge).
The method to tension the bolts involved using hydraulic jacks to apply tension, then running the bolts down hand tight. Then a measurement using a micrometer was taken and compared with the original length of the bolt. If it was out of spec the process was repeated.
torque gives a good approximation of the tension required. steel is actually a "plastic" material. There is a range of tension that a bolt will work in below where it yields. Torque figures are typically conservative.
BTW Ive been building 911 engines for about 20 yrs, I always use a torque wrench.
