NSX2B - While you can certainly make sure a wheel isn't going to fall off by "checking" it w/ a torque wrench, you can't accurately determine if it's fully torqued because there is a difference between static and kinetic coefficients of friction.
Hrant - I am not missing the point on what a margin of error is, I'm trying to explain that that it's not being used the way you seem to think it is. You would be right if they really meant it as a true per-measurement deviation error, but it is a margin of calibration error. In other words, it is an <a href="http://traffic.ce.gatech.edu/nchrp2045/v1chapter4_files/image004.jpg">accuracy</a> margin of error, not a <a href="http://traffic.ce.gatech.edu/nchrp2045/v1chapter4_files/image004.jpg">precision</a> margin of error.
As for the difference in price, the $89 wrench will typically outlast the $19 wrench. It also typically comes with a padded hard case, whereas the cheap ones often do not. The cheaper ones probably generally have a greater precision margin of error as well. But for someone who uses them infrequently to torque lug nuts, those things may not be of concern to the buyer. That's why they make both.
Torque is an imperfect measurement to use for this purpose anyway, but it is "good enough" if done properly, and the only practical one to do in your garage. Real tension analysis equipment is completely impractical for this application.
I'm done with this thread. Everyone can do whatever they want. I was just trying to help people who might be interested in doing things the right way.
