Brian,
seems like you are using the pull scale in the correct manner, as long as the bolt you are attaching to, isn't sticking out to far from the rotor plane, since that would add a small lever arm ( but probably insignificant). The lever arm you are supposed to be using is the offset of the rotor axis to the yaw pivot x force exerted by the wind on the rotor. So you should be pulling in the direction of the axle axis.
I contributed to an older thread on the question of how much thrust is exerted by the wind on a rotor that is not furled. There were several equations which contradicted each other, including one that was higher than that of the wind itself.
If I can summarize the various comments, the efficiency of the turbine at any particular WS, relates to what % of the total avail power in that WS for a specific diameter rotor is then exerted along the axis towards the tower.
As an example, for a 16 foot mill like DaveB's, in a 25mph WS, the total power in the wind is 326 # force, but if running at Cp=0.35, then about 100# is pushing backwards
Total force = 0.0026 X (Area) X (wind velocity ^2)
If the axis is offset, then a yaw force pushes against the tail, wherein the mounting angle and weight of that tail gives the reaction force you are measuring statically in your test setup.
Changing subjects slightly:
There was also an earlier thread by you late 2007 asking about blade designs and sheet metal construction ..what was your final design as far as material, diameter, pitch, profile, taper, twist??
Stew Corman from sunny Endicott