1 - ground, 2 hot, 3 cold...
The terminology changed over time to be more explicit. The Audio Engineering Society (AES) maintains standards to help interoperability between different vendors, and the current standard for XLR style connectors is AES-14-1992-S2014; that document lists pin 1 as being "Screen (shield)."
In the past you did often see pin 1 referred to as the "ground" pin, but that is ambiguous and does not make clear that it should be connected to chassis (which is often connected to safety earth/power ground) and not the circuit power supply reference point (often referred to as "circuit ground").
then part of the ground that goes to 1 (half of the copper strands), is soldered to the metal tab.
That is sometimes done to enforce the "pin 1 connects to chassis at the end device" requirement for proper shielding, but is counter to AES-51-1 which specifies that the shield should connect only to pin 1 of the cable connector and not to the connector shell.
The reason for that requirement is that if the shield connects to the shell, if the shell contacts building metal (in a commercial building all exposed metal is required to be tied to the safety earth system) there can be power supply related currents flowing unexpectedly through the cable shield to the connected equipment. That can cause noise problems in susceptible equipment, and in general can be a nuisance because the noise behavior changes depending on where the cables are lying.
That noise susceptibility from cables contacting the building safety earth would be a 60Hz hum in North America (50Hz in much of the rest of the world), and is not really the same thing as the white noise you are describing.
I ended up cutting the shield (4th solder point) on both ends of the xlr cable and reconnected it. The noise went away, and the microphone had tons of gain and was very sensitive, with only a little boost from the pre-amp
The description of the mic becoming more sensitive sounds like possibly some of the shield strands had come loose and were touching either pin 2 or pin 3 in addition to pin 1 and the shell tab, i.e. shorting one of the signal pins. Did you notice if that could have been the situation?
Does the microphone require phantom power? For a dynamic microphone shorting one pin to shield would make it more susceptible to noise pickup, but if it reduces the sensitivity at all should not be by more than 6dB. The large decrease in sensitivity you describe I would expect if a phantom powered mic had one leg of the phantom power shorted to shield.