The bolt is a tension spring and it has to stretch under load to provide clamping force. The stretch must not be large enough to permanently deform the metal. Otherwise the bolt cannot return to its original length after stretching and it provides insufficient clamping force. The limits in the chart are likely the limits on stretch values that will not permanently deform the bolt and they include a safety factor.
Bo,
This is a really complex subject, with lots of nuance regarding total load Vs fastener diameter, fastener material, serviceability, thermal loads, part re-use, and other complex issues.
Interested Observer is correct about deformation/elongation. In a nutshell, if he was not, torque to yield fasteners would not work. He is also correct about fastener manufacturers being very conservative in their recommendations. And torque to yield does work well. OEM's would not specify its' use if it did not. The basic concept is that a fastener in a "slight" yield condition is better than an "improperly" installed fastener. Fasteners, of course,
MUST be replaced when servicing. Not an issue for OEM servicing, but dicey when the "quality" of the service tech is at issue. I know a "Ferrari mechanic" (self labeled) who refuses to own a torque wrench,
because, he knows "how tight to make things". The engines he works on suffer from several issues . . . . . . let your imagination run wild.
And the strength of the casting, where studs and/or bolts engage, is also an important factor. Obviously, the best bolt can not compensate for any clamp load lost by engagement in a low strength casting.
F/b