I honestly don't use a torque wrench very often on my bikes. I used to use them a lot more on cars. But I decided to see how close "theoretically" they were. I got a digital scale with a metal handle and hook and tested a weight against a kitchen scale value to make sure the digital scale wasn't way off.

I have a:
1/4" 10-200 in/lb
3/8" 100-750 in/lb
1/2" 10-150 ft/lb.

The main difference I did was to slightly expand the hook to slide over the wrench handles. This way I can find a shoulder toward the end where the handle knurling stops to use as a more solid location to test the wrenches. This means it is not testing at 6" or 12", it is at whatever the distance to that point. If it measures 10-7/8" for 100 in/lb setting I use that. Turn it to 150 in/lb setting and if it is 10-13/16" I use that distance. There is nothing special about whole numbers, it is just simple math.

What this did was give me a rough range of variances over 8/9 torque settings on each torque wrench. The object is not necessarily to calibrate the wrenches: which I found was tough when an end cap would not unscrew, but to get a feel for where the numbers ended up. Since I didn't want to clamp of the micrometer handle with more force than my bike workstand to loosen the end cap, I made index cards of the different in/lb or ft/lb variances at each setting and put one in each case. I used the cards because 6 months from now I won't remember a thing.

I did a comparative testing between wrenches on a metal carriage bolt on a dolly to make sure I didn't do something stupid and that the adjusting for variances gave me about the same click.

At the end of the day, I will still use a basic allen wrench to secure bottle cases, tighten seat posts, clamp cables, etc. And if I had carbon I'd probably spend the money on Nm T-handles.

John