You simulate at the angle that would yield failure at the lowest loading (intuition or trial and error), then set your minimum safety factor for your listed WLL from this number. That angle will probably be perpendicular to the plane of the eyelet. Are these made from A36 steel plate? Don't forget a HCF simulation also, for those who like to get stuck often. It may cover you if you do this and have a 3rd party load test your findings. Clearly stamp the WLL and breaking strength on them. It would be good to know what an average peak force is for a recovery using a strap or cable (not a yankum style rope). Could use a load cell and test if you had one big enough. As for the bolts you cant control what people use but you could specify a minimum size and grade. For simulation just affix one side of the mounting holes, could use the edge. Or could affix the surface of the plate that would touch frame and assume no separation between the frame and your component. Cool machining and Nice part!No I have not but similar styles made from 3/8" have a 15,000 kg rating, mine being 1/2" should have higher yet similar numbers. The next run I may do a simulation on.
Some factors such as Hardware and pull angle, and pull type would be interesting to see numbers on.