To see if you have a chance, read the resistance between the spade terminal and the case (ground). You say the voltage there was ground (zero). Since your gauge was open circuit, that would be true as long as the sender resistance is finite (not infinite). If the resistance is zero, the internal points are closed permanently (consistent with your permanent >full scale reading). If the resistance is not zero (say 100 ohms) but not infinite, then it has a chance to be working.
Here’s the proper complete way to bench test the sender. You can improvise if you lack the complete equipment setup.
Connect the + side a variable voltage power supply (at least 9-15 volts) to the + side of a working gauge. Connect a wire between the sender side of the gauge and the sender spade terminal. Clip onto the case of the sender (pipe thread is convenient) and connect that wire to the - side of the power supply. Set the power supply to 13.5 volts or so. The gauge should read near zero. A voltmeter across the sender should show pulses (repeating every second or so) in voltage between 12V and something less than 12V, say 6V. I assume you don’t have an oscilloscope, the correct way to do it.
Next, you need a supply of compressed air or other gas at variable pressure–from a tank or from your shop compressor. Connect the air to the gauge. It should read in proportion to the pressure.
Last, set the pressure for, say, 40PSI indicated on the gauge (power supply still set to 13.5V). Vary the supply voltage. The pressure reading should stay fairly stable–not vary as much as the voltage does. For example, going from 9 to 15 volts should NOT change the pressure reading proportionately.
I’ve tested perhaps a dozen senders…usually they fail by not switching or switching erratically. In other words, they either work or they don’t. The method above allows you to cherry pick among working senders, though. IMHO. Hope I’ve remembered this correctly–I’m at the keys in the refrigerator age.