The unwanted dot at the Unbalanced Signal (+) input seems to be an artifact of the schematic drawing package. Take a look at the 5.1K input resistor on pin 6. It too has a dot at the 90 degree turn where the line connects to Unbalanced Signal Gnd input. There are unnecessary dots all over the diagram. If the vertical line had been drawn over the left by one grid unit, the dot would still be there, but it would not be at the intersection of the lines crossing each other.
Other than that, the circuit looks to me like a garden variety unbalanced-to-balanced signal converter optimized for a high RF environment. I built several of them virtually identical to this one, but I ultimately decided that it was easier, cheaper, and quicker to use a transformer. Plus, I like the sound of transformers.
What is interesting is the designation of Out Pin 2 as Hot and Out Pin 3 as Cold. The signal on Pin 2 is inverted or out of phase from the Unbalanced Signal (+). Moreover, the circuit is really a balanced-to-balanced converter. The inputs are totally balanced. In that application, except for the RFI filtering, you could use two pieces of wire to do the same thing.
It would be very easy to debate the need for or value of particular components in the circuit. For example, the 47uF caps are not necessary. The power supply is bipolar, and the op amps are operating at unit gain, so there will not be a DC offset at the outputs. I would argue that the caps are there to protect the circuit from an DC that might appear from the circuit following this one. The 5.1K resistors from the 47uF to ground are to keep the cap from charging, and the 120 ohm resistor is to prevent the 5532 from self-destructing if the circuit is connected to an improper load.
What is the source of the schematic? What is the intended application for this circuit?