Depends on the skill of the welder. There is nothing inherent in short circuit transfer that makes it prone to bend problems other than the arc energy being imposed on the puddle contributing to a lack of fusion. If welder skill provides good fusion then a short circuit transfer weld is every bit as ductile and strong as any other transfer, given minor variances due to the same thing, low arc energy.
A 70S-6 deposit is a 70S-6 deposit.
Yes, I have seen welders pass the bend tests using GMAW(S), myself included. Which just proves anyone can weld if they have proper instruction.
The "trick" is not to depend on the electrode melting and filling the groove and assuming the molten weld puddle has sufficient superheat to fuse to the bevel face(s). The welder has to limited the weave width and direct the electrode to the point where fusion is desired. In other words, most welders will oscillate the electrode from side to side, but they do not direct the arc directly toward the bevel face. The oscillation will produce a weld bead with the proper width and volume of weld, but the molten metal will not have sufficient superheat to fuse to the bevel face, instead, the groove acts like a metal mold and the weld metal simply casts to the shape of the groove. In extreme cases, the original groove face will still be evident when the bend samples fail.
I've taken fillet tests, 3/8 inch and 1 inch grooved plate tests and passed all three in multiple positions. The metal has to be ground to bright metal, i.e., no mill scale, and the proper manipulation must be used to ensure the molten metal fuses to the bevel face properly. Every pass has to be completely cleaned and wire brushed between passes.
This is where proper training is important to teach the welder the proper techniques.
Best regards - Al