With regards the sketch by TTM, am I correct in thinking that if the lengths (at centres) of the converted ring spanner and the torque wrench are the same, and right angles are maintained, the torque figure is the same as using a torque wrench straight on a bolt/nut, and the only thing that alters the torque imparted is the flex in the extension and anything else that flexes.

If this supposition holds up, (if it doesn't, please enlighten me) how does one make allowances for any lost toque due to flex?

Also, what if (as in my case) the torque wrench is, say, for instance 153% the length of the converted spanner? A really simple formula would be appreciated (I didn't do maths at school, I did rolling fags in the bike shed)

Glenn