The plate problems showed up when the well exploded with our equipment on top of it. Imagine two parallel plates separated by a distance of 48 inches and four legs welded between them at the corners. Load this assembly in tension at the centers of the plates and you have an approximation of what we had.
The design was "homemade" and we found that there were others in service by our field personnel. After loads of investigation, we determined that rolled shapes, especially plates, should never be loaded in tension in the through-thickness direction. There are rolled plates that you can purchase that have improved through-thickness properties, but I wouldn't trust these based on what we found in our research. However, this all relates to rolled plate. You can purchase plate which comes forged from the mill, but that's not the same thing as rolled stock. I've never heard of a problem with forged plate, but I also have only used it once (so far, the part has worked fine).
However, that being said -- if you're talking about a lug, i.e. a pad eye, then I'm mostly unconcerned about your application. In fact, I would even skip the ultrasonic testing for your plate stock and I wouldn't even make a recommendation to use the round/bar stock, assuming that all of the loading is in the plane of the plate. For this type of application, I use common rolled stock all of the time.
Additionally, on pad eyes for skids, spreader bars, guying anchor points and other parts, I have never made a consideration for grain direction. Someone may differ with me on this, but I have also never read any requirement to take it into consideration in any specification that I have come across.
Now another point -- if you are talking about using plate and cutting it with a CNC, even if you are using water jet cutter or something equally fine, remember that you can cut the shape with the CNC, but ALWAYS ALWAYS ALWAYS drill the eye. It doesn't have to be reamed or bored, (although you can) but NEVER cut the hole with a CNC, no matter how fine the process, unless it's a CNC mill. The reason is because with a plasma cutter, flame cutter, water jet cutter and even EDM wire, they will leave cutting marks in the wrong direction, i.e. parallel to the axis of the hole, which WILL make a difference when it comes to fatigue.
So still assuming that you're using a pad eye and a shackle:
1) It is best to have a tight-fitting pin in the eye, so mill/drill the hole diameter no larger than 6% greater than the nominal diameter of the shackle pin. Try to stay as close to 6% as you can in order to prevent contaminants like paint and rust from allowing for free rotation of the shackle in the eye. This number comes from DNV 2.7-1 Rules for Offshore Containers. If you use this limit, then you will be safe as far as contact stresses go. According to the spec, allowable contact stress is twice yield. (yes that's a 0.5:1 SF for contact. But only for contact)
2) Make sure that the thickness of the plate is at least 75% of the width of the jaw of the shackle. (also DNV 2.7-1) I try to make sure that it is about 90% of the width of the jaw so that nobody can "accidentally" use the wrong size shackle in the pad eye. The next shackle smaller won't fit over the plate and the next shackle larger can't fit the pin into the hole.
3) Only use bolt-type shackles with secondary cotter pin retention (see Crosby model G-2130) and don't use screw-type shackles as these have an annoying tendency of backing themselves out in service. My organization has completely banished the screw-type shackles and personally, I don't think that the screw-type shackles should even be on the market.
4) With a 4:1 safety factor, I wouldn't waste much time worrying about fatigue. Just make sure you have a reasonable inspection regime.
Are you talking about using shackles and pad eyes? If so, then I don't think it's a terrible idea at all.
Engineering is not the science behind building. It is the science behind not building.
