6 thou? How could such a gross oversize get through? I'd expect an interference fit and 6 thou would have been obvious. Even close fit is on 5-15 ten thou and 6 thou should have been clearly too loose when installing.
Yes it would be obvious if the hole in ONE piece was perfectly in line with the hole in the other piece. But my **GUESS **is that the hole in the pickle fork was drilled in a jig, and the hole in the Fail safe strap was drilled/enlarged on final assembly, and most probably never reamed as an assembly. Hand drilling a hole can easily result in an out of round hole or slightly oversize. An we do not know the exact assembly process. Or IF interference holes/fasteners were specified. trying to hammer in an supposed light interference hole during assembly creates other problems. For example one could coldwork a hole, and not ream i giving an elongated hole which partially defeats the whole purpose. There is of course the incorrect ' logic ' that reaming a coldworked hole destroys the advantage of prestress around the hole.
Properly done, with correct size mandrel, the stressed area of significance typically exists for one radius external to the origional hole eg start with a 1/4 nominal hole, radius = 1/8 inch. = giving a radial stressed area equal to about 1/2 inch diameter. And usually due to how the aluminum part was processed/rolled, the resulting hole will be slightly elongated ( ellipitical ) by as much as a few thousandths, again depending on material and hole size. Thus to mmaintain a proper interference fit, a follow up ream is required. The preceeding is an overly simplIfied explantion of course.
Sort of a side note- if you want to raise the hair on your friendly stress person, talk about vertical scratches in a structural hole... which then leads to trying to make a sleeve with a spiral opening. Which made early attempts to make a practical sleeve a real problem. After much soul searching, a simple vertical split in the coldworking sleeve was tried, worked great. But still the issue was a major discussion of what about- what if.... etc. So a significant test program was initiated in 60's-70's , with groups of holes and individual holes coldworked with various sleeve split orientations to see if eventual failure was along the vertical split opening or somehow related. Bottom line- end result- no significant relationship re sleeve orientation and eventual overload- fatigue failure.