In a receiver this is probably not true. The first thing you need to do is convert the output to single ended to drive the amplifier. So balanced is wasted. True balanced amplifiers designs exist, but are little more than an excuse for overly complicated designs with dubious actual advantages in performance, and some very real down sides.
For a processor, if there is a need for longer runs to the amplifiers, something that may result in real world noise and ground loop issues, balanced has clear benefits. But if you are simply stacking the processor and amplifiers in the same rack, it is quite possible that balanced is a net loss in overall performance. Even with balanced outputs, driving the two sides of the balanced outputs as a pair of individual balanced signals all the way from the DAC outputs is also potentially complicated and fraught with issues. You need dual reconstruction filters, and you need to manage the RF energy in the system with larger path loops resulting from the more complicated design. And you must still provide galvanic isolation if you want the performance benefits. Given the essentially perfect performance seen in quite mundane and cheap systems where little more than careful adherence to layout, grounding and decoupling are all that is needed, the complexity, cost, and potential performance issues would suggest staying away from dual differential balanced for any reason other than marketing. Maybe if you have a true cost no object design, and need to justify the resulting eye watering cost, this is a way to go. But it won't sound any better. Just get you bragging rights.