What evolutional benefit is that?
They probably didn’t need their arms for how they hunted. Same as modern birds. Think of how an eagle kills and then eats its prey.
I’m speaking more to them grabbing the prey with their legs and just holding it to death. They also rip and peck with their beak as they hold the unfortunate snack. The same would probably be true for dinos with short arms minus the beak thing but they would use teeth instead.
Go look at an ostrich skeleton, and you’ll quickly realize that the arms on a t-rex skeleton have been posed backwards from what they should be. They’ve traditionally been posed based on the assumption that the skeletons were related to lizards. In reality, they’re avian, (we’ve found plenty of evidence that they had feathers, for instance,) and the tiny arms easily could’ve been wings like an ostrich. An ostrich skeleton also has the distinctive tiny arms like a t-rex, but they’re rotated 180 degrees to work as wings instead.
Their arms are small, but beyond that there’s basically nothing similar between them and an ostrich’s wing. The muscular anchor points are not similar at all to winged creatures, who require significant musculo-skeletal connection to the breastbone even in mostly vestigial wings. You can see this in the ostrich skeleton as the large “blob” of bone in the middle of the rib cage. There is nothing similar in the T-Rex. Even more of a problem with this theory is that the T-Rex’s popularity is in large part due to the fact that we’ve discovered a fairly large number of T-Rex fossils in good condition and not substantially disturbed… It’s why we have famous models like “Sue” and “Black Beauty” that make such good displays in natural history museums. Unless you’re proposing that a dozen different skeletons from several different regions with different ages all had bones shift after death to end up in the same position…
Our knowledge of what dinosaurs looked like is not perfect, but we’ve also come a very long way from the Magdeburg Unicorn or horned Iguanodons of the 1800s. Paleontology has largely moved past “puzzle piece” biology, where things are just haphazardly thrown together because they kinda look like they fit. There’s comparison to other species - not just reptiles- to see what are comparable modern equivalents or to other contemporary animals. There’s kinematics and musculature considered. Unless some fossil discovery is made that completely upends the evidence we have now, at least in the case of skeletal articulation of well-known and well-studied species like T-Rex, we can be reasonably confident that we’ve got it pretty close when it comes to what their skeletons looked like.